<%BANNER%>

Destination Message Design on Travel and Tourism Information Web Sites

Permanent Link: http://ufdc.ufl.edu/UFE0015748/00001

Material Information

Title: Destination Message Design on Travel and Tourism Information Web Sites
Physical Description: 1 online resource (152 p.)
Language: english
Creator: Lane, Charles W
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: cvb, design, destination, instructional, message, tourism
Tourism, Recreation, and Sport Management -- Dissertations, Academic -- UF
Genre: Health and Human Performance thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The rapid expansion of the Internet has changed the way people communicate. The Internet is the primary source of information in many aspects of our lives, including travel planning. Tourists expect to be able to gather online the information they need to plan their travel. Convention and Visitors Bureaus (CVBs) exist to provide travelers with information about destinations. Unlike the sites of airlines, hotels, and attractions, the primary mission of most CVB web sites is not sales. Instead, a CVB web site informs users about the location it represents. Literature from instructional message design and learning theory research supports the notion that an effective informational web site, such as that of a CVB, should employ supportive message design elements to reduce the cognitive load of the user to enhance information acquisition and processing. Evaluating web sites according to instructional message design principles is a method of appraising web site effectiveness in enhancing information acquisition and processing; however, in the commercial sector, such techniques have not typically been applied. The purpose of this study was to examine the use of supportive message design elements on CVB web sites worldwide. A content analysis of 588 sites was conducted to determine which elements were used and to what extent. Statistical analyses were conducted to determine if there were differences among regions of the world in the use of effective design principles. Results indicated that many message design practices were nearly ubiquitous on CVB web sites. Web sites were more effective in their use of visual display elements than in their use of content presentation elements. Message design elements can be reliably grouped into four dimensions: graphics presentation, text presentation, graphics/text support, and visual text. Regional differences were found on nine message design elements and one underlying factor. Expert web developers considered functionality and aesthetics to be the most essential features, revealing a lack of understanding in the field of the importance of the design of message content delivery. Results are discussed in the context of the theoretical framework and previous research, and recommendations for practice and for future research are provided.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Charles W Lane.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Holland, Stephen.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0015748:00001

Permanent Link: http://ufdc.ufl.edu/UFE0015748/00001

Material Information

Title: Destination Message Design on Travel and Tourism Information Web Sites
Physical Description: 1 online resource (152 p.)
Language: english
Creator: Lane, Charles W
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: cvb, design, destination, instructional, message, tourism
Tourism, Recreation, and Sport Management -- Dissertations, Academic -- UF
Genre: Health and Human Performance thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The rapid expansion of the Internet has changed the way people communicate. The Internet is the primary source of information in many aspects of our lives, including travel planning. Tourists expect to be able to gather online the information they need to plan their travel. Convention and Visitors Bureaus (CVBs) exist to provide travelers with information about destinations. Unlike the sites of airlines, hotels, and attractions, the primary mission of most CVB web sites is not sales. Instead, a CVB web site informs users about the location it represents. Literature from instructional message design and learning theory research supports the notion that an effective informational web site, such as that of a CVB, should employ supportive message design elements to reduce the cognitive load of the user to enhance information acquisition and processing. Evaluating web sites according to instructional message design principles is a method of appraising web site effectiveness in enhancing information acquisition and processing; however, in the commercial sector, such techniques have not typically been applied. The purpose of this study was to examine the use of supportive message design elements on CVB web sites worldwide. A content analysis of 588 sites was conducted to determine which elements were used and to what extent. Statistical analyses were conducted to determine if there were differences among regions of the world in the use of effective design principles. Results indicated that many message design practices were nearly ubiquitous on CVB web sites. Web sites were more effective in their use of visual display elements than in their use of content presentation elements. Message design elements can be reliably grouped into four dimensions: graphics presentation, text presentation, graphics/text support, and visual text. Regional differences were found on nine message design elements and one underlying factor. Expert web developers considered functionality and aesthetics to be the most essential features, revealing a lack of understanding in the field of the importance of the design of message content delivery. Results are discussed in the context of the theoretical framework and previous research, and recommendations for practice and for future research are provided.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Charles W Lane.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Holland, Stephen.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0015748:00001


This item has the following downloads:


Full Text












DESTINATION MESSAGE DESIGN
ON TRAVEL AND TOURISM INFORMATION WEB SITES










By

CHARLES W. LANE


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2007




























Copyright 2007

by

Charles W. Lane














ACKNOWLEDGMENTS


This study was conducted with the assistance and support of numerous

individuals. My sincerest gratitude is extended to all those who contributed.

My sincere appreciation goes to Dr. Stephen Holland, my committee chair,

who has provided me with so much support. My heartfelt thanks also goes to Dr.

Heather Gibson, who provided thoughtful and relevant feedback and whose

constant encouragement has kept me going. Thanks to Dr. Lori Pennington-Gray

for adding so much to my doctoral program. The expertise of my external

committee member, Dr. Colleen Swain, has been invaluable throughout my

program.

I am deeply indebted to Dr. Brijesh Thapa. Without his support this project

could not have been completed.

I would also like to express my thanks to my fellow doctoral students,

especially Jung-Eun Kim, Louisa Meyer, and Lisa Pennisi, who have made this

experience enjoyable and provided much needed support.

Valuable input also came from Dr. Benjamin Lok and Dr. Lee Mullally.

The hard work of Michell York and Mario Klemmer contributed significantly to this

study, and their efforts are deeply appreciated.

Special thanks are also due to my master's guru Dr. Michael Kane,

without whom my graduate education would not have been possible.

Finally, and most importantly, the love, support, and encouragement that I

have received from my family have made this endeavor not only possible, but

also worthwhile.















TABLE OF CONTENTS

ACKNOW LEDGMENTS .......... ............... ....... ..... ............... iii

LIST OF TABLES ................ ....... ......... ......... vi

LIST OF FIGURES ........................................... .........viii

ABSTRACT ................ .. ......... ................. ix


CHAPTER

1 INTRODUCTION ....... ......... ...............................1

Tourism and the Web ....... ........... ............. ... .................4
Instructional Message Design ............. ......................... 11
Theoretical Fram ew ork ............. .................................................... 12
Purpose of the Study ......................................... 13

2 REVIEW OF THE LITERATURE ........ ........ ............................... .15

The World Wide Web ................ .. ............. ......... .... ............ 15
Technology in Tourism .................................................... 23
Internet W eb Site Evaluation .................................... ................... 31
Instructional Message Design........... ... ...... .. ......... ............ 33
Cognitive Load Theory ................................................ ..................... 50
Application of Cognitive Load Theory and Instructional Message
Design Principles to Tourism Information Search............................54

3 M ETHO DS .............. ..... ....................... .............. .... ............ 58

Methods: Research Questions 1-4 ................................ ................... 59
Methods: Research Question 5 .......... ......................................68

4 RESULTS ............... ... .............. ....................... 72

Results of Data Analysis................ ......... ..... .......... .. ................78
Sum mary .................................................96




iv












5 D IS C U S S IO N ...................................................................... 9 8

Discussion of the Results .......................... ........ ............ ............... 98
Limitations and Delimitations of the Study .................................... 16
Implications for Practice .......... ........................... ............ 117
Implications for Future Research ...... .... ......... .......................... 119


APPENDIX

A INSTRUCTIONAL MESSAGE DESIGN ELEMENTS ASSESSED
ON CVB W EB SITES ...... .... ............ .................... .............. 122

B DELPHI PANEL WEB PAGE ELEMENTS IMPORTANCE
C H E C K LIST ...... .........................._..... .. ................ ... 126


R E F E R E N C E S ............................................ ................ 13 1

BIO G RA PH ICA L S KETC H ...................... .. ............................ ............... 142









LIST OF TABLES


Table page

2-1 Support in the literature for web site content elements ................20

2-2 W eb site evolution by generation ................................................. 34

3-1 Sample from CVB world regions ................ ............. ............... 60

3-2 Rater summary scores on primary component ..............................65

4-1 Response variability of the Likert-type items.................................. 73

4-2 Response location of the survey items ....................... ......... ......74

4-3 Item discrimination of Likert items (alpha=.862)............................. 76

4-4 Item analysis composite...................................... 77

4-5 Prevalence of instructional message design elements on CVB
web sites ............... ....... ....... ......... ...............79

4-6 Ratings of message design element use on CVB web sites............81

4-7 Factor analysis and reliability of the message design variables....... 85

4-8 Message design elements exhibiting significant differences
betw een reg ions.......................................................... 87

4-9 Graphic presentation significant difference by region ...................88

4-10 Broadband download times by world region ................. ........ 89

4-11 Flesch-Kincaid grade level readability scores by region ................. 89

4-12 Leg ibility of text by reg ion........................................... ..................... 90

4-13 Color vision impairment mean scores for text by region .................90

4-14 Legibility of headings by region........................................................ 91

4-15 Significant difference between text and heading color difference ....91

4-16 Significant difference between text and heading brightness
difference ... .. .................................... ... ............... 91

4-17 Color vision impairment mean scores for headings by region..........92









4-18 Text / headings difference: Protanopia ................. .......... ......... 92

4-19 Text / headings difference: Deuteranopia .................................... 93

4-20 Text / headings difference: Tritanopia......... ....... .................... 93

4-21 Delphi panel importance ratings of message design elements........94









LIST OF FIGURES


Figure page

1-1 Assael's (1984) model of consumer information acquisition and
p ro cessing ............................... ................ . ........... 8

1-2 The relationship supportive message design, cognitive load, and
information processing............................... ...................... 13

2-1 The role of supportive message design in the facilitation of
consumer information acquisition and processing .......... ......55

4-1 Scree plot of factor loadings................................ ..................... 83














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

DESTINATION MESSAGE DESIGN
ON TRAVEL AND TOURISM INFORMATION WEB SITES

By

Charles W. Lane

August 2007

Chair: Stephen Holland
Major: Health and Human Performance


The rapid expansion of the Internet has changed the way people

communicate. The Internet is the primary source of information in many aspects

of our lives, including travel planning. Tourists expect to be able to gather online

the information they need to plan their travel.

Convention and Visitors Bureaus (CVBs) exist to provide travelers with

information about destinations. Unlike the sites of airlines, hotels, and

attractions, the primary mission of most CVB web sites is not sales. Instead, a

CVB web site informs users about the location it represents.

Literature from instructional message design and learning theory research

supports the notion that an effective informational web site, such as that of a

CVB, should employ supportive message design elements to reduce the

cognitive load of the user to enhance information acquisition and processing.









Evaluating web sites according to instructional message design principles is a

method of appraising web site effectiveness in enhancing information acquisition

and processing; however, in the commercial sector, such techniques have not

typically been applied.

The purpose of this study was to examine the use of supportive message

design elements on CVB web sites worldwide. A content analysis of 588 sites

was conducted to determine which elements were used and to what extent.

Statistical analyses were conducted to determine if there were differences among

regions of the world in the use of effective design principles.

Results indicated that many message design practices were nearly

ubiquitous on CVB web sites. Web sites were more effective in their use of

visual display elements than in their use of content presentation elements.

Message design elements can be reliably grouped into four dimensions:

graphics presentation, text presentation, graphics/text support, and visual text.

Regional differences were found on nine message design elements and one

underlying factor. Expert web developers considered functionality and aesthetics

to be the most essential features, revealing a lack of understanding in the field of

the importance of the design of message content delivery. Results are discussed

in the context of the theoretical framework and previous research, and

recommendations for practice and for future research are provided.














CHAPTER 1
INTRODUCTION

Effective communication is critical in modern society. We are living in

what is widely known as the "Information Age"-a time in which information and

the manner in which information is communicated is more important than ever

before. Increasingly, communication of information occurs via computer

technology. In fact, according to Naisbitt (1984), "computer technology is to the

information age what mechanization was to the industrial revolution" (p. 28). One

of the most pervasive uses of computers is to access the Internet.

In its relatively short lifespan, the World Wide Web has grown to a

massive size. The Internet is a borderless system of computer networks that

allows any computer to communicate with any or all other computers on the

network. It is, simultaneously, a medium for human-computer interaction, a

vehicle for information dissemination, and a forum for the sale of goods and

services. Although these functions all occur without respect to geographic

placement, the diffusion of the Internet itself has been characterized by sharp

disparities in technological development and accessibility (Kiiski & Pohjola,

2002).

The Internet had its genesis in the United States as a tool used largely by

scientific researchers (Rosen & Purinton, 2002). Even at its inception, the

Internet was a cluster of related technologies where accessibility depended on









both hardware and infrastructure: a computer, a remote host, and connective

telephone lines (Kiiski & Pohjola, 2002). The global diffusion of the Internet has

been achieved relatively quickly, but not with the uniformity that is commonly

believed. The primary factors that influence the transfer of technologies across

nations are the characteristics of the technology, cultural variation, societal

differences, and the absorptive capacity of the recipient region. As a result, the

process of technological transfer is not linear (Kedia & Bhagat, 1988). The

degree of availability of Internet hardware and infrastructure has been influenced

by the ambivalence of governments and the corresponding irregularities in

government policies (Goodman, Burkhart, Foster, Press, Tan, & Woodard, 1998).

For developed countries, the diffusion of the Internet and related

technologies has been supported by governments and large corporations

through the creation of research and academic networks that serve as the

backbone of the enterprise. For most other countries, these types of national

technological backbones are beyond the scope of their financial resources, and,

consequently, their national priorities (Goodman, Press, Ruth, & Rutkowski,

1994).

By September 2004, there were an estimated 287.5 million English

language web users online and an additional 516.7 million non-English language

web users (Global Reach, 2004). Current statistics estimating the total number of

web pages on the Internet have generally been offered alongside a warning of

their unreliability. The majority of homes in America now have at least one

computer, and 64% of all Americans have used the Internet in the past year









(Rosen & Purinton, 2002). The Internet has become the primary technological

medium for delivering online content, such as the information, features, or

services found on web sites (Huizingh, 2000), in addition to instructional material

(Montilva, Sandia, & Barrios, 2002).

Internet connectivity is strongly correlated to GDP per capital across world

regions, where people in wealthy, developed areas enjoy a higher rate of access

than those in undeveloped areas. Since information technology is thought to be a

driver of economic growth, the inequalities in the diffusion of Internet technology

may be a contributing factor to the disparity of income potentials between the

richer and poorer world regions (Kiiski & Pohjola, 2002). In addition, technology

can contribute to sustained inequality due to its uneven distribution throughout

areas of the world. Given the importance of the Internet, the level of diffusion

can influence whether an area can sustain its place in the global economy

(Harhittai, 1999), and the potential for impact on the tourism industry is great

(Frew, 2000). Sigala, Airey, Jones, and Lockwood (2000) explained, however,

that due to the nature of the tourism industry, characterized by multiple small

businesses, the use of technology by tourism professionals has been slow to

diffuse to some areas.

There is widespread agreement that the Internet is evolving as a dynamic

electronic marketplace at an unprecedented pace (Rayman-Bacchus & Molina,

2001). Businesses, then, need to orient their marketing strategies to address

wider diffusion of the potential consumer base (Rai, Ravichandran, & Samaddar,

1998). Over the past few years, the Internet has undergone a wave of









commercialization where attention has been focused on the use of the global

information infrastructure to support commodity and other commercial services

(Leiner et al., 2003). Projections of continued growth serve to inform providers of

Internet products and services of the potential worldwide market.

Because the Internet is a phenomenon that grew very quickly, the rush to

generate a web presence prevented many web content providers from

researching the needs and wants of web consumers (Hodkinson & Kiel, 2003)

prior to going online. Consequently, web page designers, where design refers to

the way that content is made available to web site visitors (Huizingh, 2000),

produced a variety of information delivery formats in a research vacuum.

Belatedly, content providers have turned to empirical research to examine the

needs and wants of web-based consumers (Hodkinson & Kiel, 2003).

Tourism and the Web

As a correlate to the growth of the Internet, there is evidence that the

classic model of the tourism product is changing from mass tourism markets

dominated by a few providers to a broader array of more specialized types of

tourism providers (Sigala et al., 2000). This has, in part, been shaped by the

evolution of technologies important to the development of the burgeoning travel

industry. The arrival of the jumbo jet in the early 1970s, along with computer-

based reservations and travel information systems, offered travelers greater

destination choice, but was still dependent on the services of various travel

agencies to enable that choice.






5


The nature of the tourism industry, sometimes characterized by multiple

small businesses, has resulted in the slow diffusion of technology and Internet

access in some areas (Sigala et al., 2000). For the tourism industry, the Internet

has also emerged as an electronic travel agent-an easy access platform that

brings travel information to consumers directly without using traditional

intermediaries like travel agents (Kroil, Kapsammer, Pr6ll, Retschitzegger, &

Wagner, 1997).

The development of the Internet and World Wide Web has provided

individual travelers with a much greater degree of autonomy in travel information

search and destination selection. At the same time, the Web has enabled small-

to medium-sized tourism providers to compete with mass tourism destinations for

individual traveler's attention (Rayman-Bacchus & Molina, 2001).

It is widely accepted that the travel industry is more likely to be affected by

technological advances than other industries, as travel has consistently placed

among the top three product categories purchased via the Internet (Weber &

Roehl, 1999). Machlis (1997) reported that the largest revenue generator among

consumers on the Internet was the travel industry. The traveler who was once

forced to rely on a travel agent can now make direct inquiries of airlines, hotels,

and destinations (Baines, 1998).

Destination marketing organizations, such as state and national tourism

offices and convention and visitor bureaus (CVBs), can provide a web presence

for many smaller tourism businesses. Convention and visitors bureaus, in

particular, seldom have a product of their own to represent, but instead act as









information brokers between tourism destinations and potential visitors (Palmer &

McCole, 2000). That information brokerage is enhanced by the maturation of the

Internet and World Wide Web, where delivery of information through high quality

message design is possible.

According to the Destination Marketing Association International (DMAI,

2006), the purpose of a CVB is to represent selected destinations to meeting

professionals, tour operators, and individual visitors. The mission of a local CVB

is to present information about a geographical area (e.g., city, state, region) to

the public, with the goal of increasing visitation to the area through tourism and

meetings.

Publicly funded bureaus, like some CVBs, have had to become

increasingly transparent with respect to expenditures of public dollars. The

shrinking of the tax base in some areas along with erosion in city and county

budgets makes a target of the revenue generated from hotel room bed taxes

(Schweitzer, 1997). Some bureaus are facing budget cuts despite their

classification as revenue producers. As a result, CVBs have responded by

initiating service improvements targeted at remaining competitive in the online

marketplace.

Adapting their traditional ways of doing business to the Internet has

become the focus of CVBs nationwide. Schweitzer (1997) explained that CVBs

now attract new business by dedicating resources to using technology as an

effective way to promote the destination. A nationwide trend among CVBs is

enhancement of web-based services to provide more comprehensive information









to various visitors and yield both advantages for consumers and benefits for the

bureaus themselves. Providing information to potential visitors is one of the

primary goals of an organization's web site (Jesitus, 1998) and central to the

primary mission of a CVB.

Since the Internet has become an integral part of daily life in the US and

elsewhere, it has developed importance to CVBs. Lake (2001) found that 93% of

travelers used the web to gather travel-related information. To reach web-savvy

travelers, maximizing the effectiveness of their web presence should be a priority

for CVBs. The primary external source for travel and tourism information is,

increasingly, a destination's web site. Travel consumers have readily accepted

this technology because it is convenient and easy to work with (Kroil et al.,

1997).

Vogt and Fesenmaier (1998) considered information search behaviors in

the travel and tourism industry, and expanded the role of tourism information

search from a primarily marketing purview into a broader communication context.

According to Vogt and Fesenmaier, it is well established that visitors to a web

site often come primarily to gather information. Tourism information seekers

primarily collect and use information for functional reasons, such as planning or

taking trips. A later decision to purchase may be greatly influenced by the

information-gathering phase of the visit. To guide potential tourists toward a

purchase decision, tourist information should appeal to a variety of search needs

to first capture the attention of potential travelers. Using Assael's (1984) model of

consumer information acquisition and processing (Figure 1-1), Vogt and










Fesenmaier explained the traveler's steps in making a tourism purchasing

decision.



Stage 1 Stage 2 Stage 3 Stage 4 Stage 5

Input Variables Search for Additional Information
Consumer:
Penm rir.lp ihit%
F'iere-onae Information
Liestyle Processing: Purcha-c
Motivations Inormation Categorize Brand and
Environmental: Acquisition Evaluate action Consume
Social Organize
Cultural Retain
Situation
Marketing-
Product External Internal
Price Sources: Sources:
Past
Place Passive Memorya
r oln Acrtivre Experience




Figure 1-1. Assael's (1984) model of consumer information acquisition and
processing.



Evaluation of Tourism Web Site Effectiveness

The technological and financial barriers to establishing a business web

presence on the Internet are few. As a result, tourism professionals have

established a web presence that encompasses diverse expectations and

formulas for success (Rayman-Bacchus & Molina, 2001). Unfortunately, the ease

of authoring web pages has created quality-control issues. The ubiquitous

nature of web site creation provides a widespread opportunity for inexperienced

information providers to create unending information delivery systems that are

difficult to use (Bevan, 1998). Web pages are often not subject to the same

quality criteria as can be found in more traditional forms of publishing (McGovern

8









& Norton, 2002). Leading travel and tourism sites, however, are visually

attractive, provide detailed and current content, and succeed at maintaining the

interest of their site visitors. The success of leading tourism sites suggests a

competitive advantage for web sites that are visually appealing and information

rich (Rayman-Bacchus & Molina, 2001). Central to the principles of instructional

message design is the presentation of information in a way that is visually

appealing and that will attract and maintain the user's attention (Fleming & Levie,

1993).

Evaluation of web site effectiveness is necessary due to the volume of

retail sales generated from Internet commerce-$74 billion in 2002 (Rosen &

Purinton, 2002) and due to the substantial costs associated with establishment

and maintenance of web sites (Tierney, 2000). The complexity of the web itself,

however, poses problems for creation of standardized measures of web site

evaluation (Gretzel, Yuan, & Fesenmaier, 2000). Tabulating direct sales from a

site is one method of evaluation, but many tourism web sites, such as those of

CVBs, do not sell directly to the consumer and cannot be measured in this

manner.

According to Tierney (2000), other common sources of data for web site

evaluation are number of hits, number of user sessions, and online feedback

from site users incorporating interactivity into a web site can enable the visitor to

relay a great deal of personal information back to the host. Another tracking

strategy employed by commercial web sites is the use of "cookies" to monitor

repeat visits, viewer preferences, and page viewings.









Statement of the Problem

Access to web information can be influenced by a variety of technical

issues, such as the speed of the Internet service provider (ISP), the settings and

capabilities of the computer involved, and the web browser being used. A

combination of these features defines the technical environment that affects the

display of information and the web site features that will ultimately be available to

the information searcher (Hodkinson & Kiel, 2003). Although web site

developers cannot control the technical environment of the end-user, successful

web site development remains, in part, contingent upon the developer's

awareness of graphic design and web usability (Bevan, 1998).

If a web site does not meet the expectations and needs of its visitors, then

it will not meet the needs of the organization represented on the site. Developing

an informational web site necessitates the inclusion of instructional, structural,

functional, and aesthetic properties (Montilva, Sandia & Barrios, 2002). When

creating web sites with the primary purpose of providing information, web

developers may use general-purpose methods similar to those used for creating

commercial web pages. The problem with that practice is that it does not

address those requirements that are intrinsic to informational web sites where

retention is the desired outcome.

To date, the discipline of instructional message design has not been

applied to the evaluation of the travel and tourism web presence as a learning

experience, as most web site developers do not consider those aspects that are

specific to the information acquisition process (Montilva, Sandia, & Barrios,









2002). Given that a primary function of CVBs is to present information to

potential visitors, travel industry professionals could benefit from understanding

ways to convey that information more effectively.

Instructional Message Design

Digital technology and interactive multimedia have had a significant impact

in the field of education, where web-based e-learning environments utilize the

power of the Internet to create a context where retention of information is

supported and enabled. Winne (1995) suggested that learning is affected by the

presentation of information, and that learning improves when the same

information is delivered in different formats (words, pictures, sounds).

A "message" is a display of text and graphics or other signs produced

expressly to modify the psychomotor, cognitive, or affective behavior of people.

"Design" refers to the process of analysis and synthesis that starts with an

instructional objective and finishes with a blueprint for that objective's realization

(Reigeluth, 1983). Messages presented as an organized set of ideas promote

understanding (Witt, 1981).

The study of instructional message design allows web designers to make

effective choices from a myriad of approaches (Fleming & Levie, 1993).

Message design principles can be best utilized as guidelines for the display of a

visual message (Pettersson, 1999). While instructional design principles and

practices appear to be critical to successful information processing and

acquisition outcomes, the specific focus of instructional message design is linking

theory and practice (Fleming & Levie, 1995). It seems logical to use effective









message design strategies to impart information to web site visitors in the field of

tourism, where technology can be used to promote message retention, enhance

communication, and provide sources of further information. Evaluation of web

sites according to instructional message design principles is a method of

appraising web site effectiveness, but, to date, this method has not been used.

Theoretical Framework

This study is based on the theory of cognitive load and the role it plays in

information acquisition and processing. Cognitive load refers to the demands on

working memory (Sweller, 1994; Sweller, Van Merrienboer, & Paas, 1998).

According to cognitive load theory, the limitation of working memory capacity is a

key issue in the learning process. To maximize retention, content should be

presented in such a way that the learner's working memory is focused on

relevant information and is not taxed by irrelevant or distracting features of

material. Information acquisition and processing should be supported.

Instructional message design principles provide guidelines for developing

effective messages. Use of supportive message design elements, such as

effective text and graphics, increases attention and motivation, which, in turn,

reduces cognitive load. Reduced cognitive load enhances the processing of the

intended message into long-term memory. Figure 1-2 illustrates these

relationships.

By providing information through supportive message design, CVBs can

enhance the effectiveness of their message delivery. Effective use of text and

graphics can promote more efficient consumer information acquisition and









processing. This study examines use of supportive message design elements on

CVB web sites.




Processing of Intended Message into
Long-Term Memory (Learning)



Reduced Cognitive Load


Attention Motivation

Supportive Message Design
Text and Graphics


Figure 1-2. The relationship supportive message design, cognitive load, and
information processing.

Purpose of the Study

The use of effective message design strategies can reduce cognitive load

on working memory and facilitate retention in long-term memory (Fleming &

Levie, 1995). It is clear that information providers in the travel and tourism

industry can benefit from using message design strategies that enable their

message be retained in the long-term memories of their web site visitors.

The purpose of this study was to examine destination message designs of

informational (CVB) web sites in travel and tourism through the prism of

instructional message design. This study assessed the use of instructional

message design elements that can promote the acquisition and retention of CVB

destination messages by web site users through reduction of cognitive load.









The goal of this study was to assess destination message design on travel

and tourism web sites and determine what effective message design practices

are used to disseminate tourism-related information across world regions. More

specifically, this study was conducted to answer the following research

questions:

1. What is the prevalence of message design elements on CVB web sites?

2. Which message design elements are CVB web sites using effectively to
deliver destination messages?

3. How many underlying instructional message design conditions that enable
the desired retention outcomes are present on CVB web sites?

4. Do regional or geographic differences exist in the use of message design
elements on CVB web sites?

5. What web page elements do web design professionals believe to be most
important for destination message delivery?

In Chapter 2, a review of related literature is presented to provide

theoretical and empirical support for the message design elements examined in

this study and their role in tourism information search. An overview of the

methods used in this study to answer these research questions is provided in

Chapter 3. The results from the study are presented in Chapter 4. Finally, in

Chapter 5, the implications of the results for consumer information acquisition

and processing theory are presented, along with recommendations for future

research and for message design strategies for CVB web sites.














CHAPTER 2
REVIEW OF THE LITERATURE

The personal computer, initially introduced in 1979, required a number of

years to become established as standard business equipment. As costs

decreased, computer ownership became more widespread. By the end of the

twentieth century, more personal computers than televisions were sold in the

United States each year (Kormier, 1999).

The World Wide Web

In its relatively short lifespan, the World Wide Web has grown

tremendously. Retail sales generated from Internet commerce were in excess of

$74 billion in 2002 (Rosen & Purinton, 2002). The growth of the Internet in sheer

number of users is changing so quickly that it is impossible to determine the

population of users at any given point in time (Kormier, 1999). For most

companies, the question is no longer whether or not to have a web presence, but

rather how to produce the best site for their particular business (Kovacs &

Rowell, 2001). Maintaining a beautifully designed, current, and informative web

site can establish a company's presence as a major player in the field and

enhance the identity of the company (Kormier, 1999). Vogt and Pennington-Gray

(2002) demonstrated the growing impact of new tourism information sources

such as the Internet, as well as decreased uses for traditional tourism marketing

approaches.









Web Site Design

A web site can be defined as a number of content elements, or pages, that

are linked together. The essential feature of hypertext design is that it provides

options for deep, nonlinear searches for information that are under the control of

web site visitors. Web designers can influence the outcomes of hypertext

searches by imposing a structure on the web site, that is, by providing links

between content elements (Huizingh, 2000).

Web sites often replicate the content and structure that mirrors the needs

or concerns of the organization rather than the needs of users. If a web site does

not meet the expectations and needs of its visitors, then it will not meet the needs

of the organization represented on the site. Web site development must be user-

centered, and the needs of the user should be paramount when maintaining or

making changes in web site design (Bevan, 1998).

As technology progresses, web sites continue to evolve, becoming faster,

richer, and more functional, with increasing numbers of visitors using them to do

a myriad of things (Motley, 2000). Web site users have more choices than ever,

and may be unwilling to waste time on sites that are slow, confusing, or do not

satisfy their needs. As a result of near-infinite choice and the ease of browsing,

web users exhibit singular impatience and demand instant gratification (Raward,

2001). Long waiting times, or what web users perceive to be long waiting times,

are not well tolerated in the online environment. The problems associated with

long waiting times are problems for web designers as well, because web sites









whose waiting times are routinely long (greater than two seconds) are visited less

frequently than others (Fui-Hoon Nah, 2004).

Jung and Butler (2000) listed some important characteristics of a

successful web site, including appearance, ease of navigation, design, strategic

partners, content, interactivity, promotion, feedback, repeat visits, value-added,

currency, and message. Gibson (2001) advises that a web site should be

interesting, fun, interactive, and user friendly, using a fair amount of graphics,

graphic icons and interactive links to develop feedback and user participation.

Many site visitors will expect two-way communication (Motley, 2000), such as

email, bulletin boards and chat rooms. Incorporating specific features into a web

site serves to improve customer service, log demographic information, open new

markets without respect to geographical constraints, increase revenues, and

reduce the number of incoming telephone calls and letters (Kormier, 1999).

The complexity of web site design involves mastery of a wide scope of

knowledge: (a) knowledge of the communication process; (b) understanding of

hardware, software, graphics, animation, video, and sound; (c) appreciating

evolving issues, characteristics, and capabilities of the Internet; and (d) analysis,

design, implementation, and maintenance of web sites (Kovacs & Rowell, 2001).

In addition, designers need to understand the goal of the web site, whether it is to

sell a product or to inform or educate a site visitor.

Instructional Web Pages

The World Wide Web is now the primary technological media for

delivering instructional material; indeed, a primary goal of many web pages is









educational in nature. When creating educational or instructional web sites, web

developers may use general-purpose methods, similar to those methods used for

creating conventional web pages. The problem with this practice is that it does

not address requirements that are intrinsic to instructional web sites. That is,

they do not incorporate those aspects that are specific to education and the

teaching-learning process. Developing an instructional web site necessitates the

inclusion of instructional, structural, functional, and aesthetic properties (Montilva,

Sandia, & Barrios, 2002).

The primary responsibility of an instructional or informational web site is to

create a place or an environment where acquisition and retention of information

can occur (Hubbard, 1998). Web pages must be well designed to provide an

environment that can enable a successful information acquisition experience.

Hyperlinks provide a form of user control over content and enable the web site

visitor to interact with the content. Hyperlinks allows users to go deeper into

subject matter using the web as a convenient electronic library (Morrison &

Guenther, 2000).

Informational web sites can be made more effective through strategies

offered from the study of instructional message design. Some researchers in

web development have offered prescriptions for web site design that are

reflective of tenets in the field of instructional message design. McFarland (1995)

states that at each step of development the developer should determine whether

text, illustrations, or icons should be used. Although the presentation of









information can be dominated by text, the option to use pictures or text or some

combination must be considered.

Web sites often contain text-based materials that were created in a

traditional print format, but have not been adapted for presentation on the web.

Web site visitors rarely read web pages word for word. Instead, they scan to find

the information they need. Web designers should make text easy to scan by

using meaningful headings, bulleted lists, and highlighted key words. Further,

large sections of text should be available for printing or downloading, as most

users will not read large amounts of text online (Bevan, 1998). Designers should

also ensure that the intended message is not skewed by the packaging of the

information. Many multimedia packages deliver more information than users can

process at one time, thus increasing cognitive load and hindering the educational

process. To protect against learner frustration, the quantity of information

presented should be limited to digestible bites (McFarland, 1995).

Color media is preferable to black-and-white (Simonson, 1984). Color

enhances communication when properly used. When color is not properly used

it may tend to confuse or even offend. Designers should use a consistent color

scheme. Soft nonintrusive background pastels and soft grays are said to provide

the best backgrounds, as the human eye may become fatigued after long

exposure to highly saturated colors (McFarland, 1995). Bevan (1998) suggests

that patterned or multicolored backgrounds can make text difficult to read.

There are two major problems inherent in web use for information

retention: (a) loss of overview as a consequence of limited screen area, and (b)









the shortened attention span of the learner due to fatigue from reading online.

Loss of overview occurs when the volume of information that can be presented

on-screen is limited compared to print media. To mitigate loss of overview, web

content should be presented in a generalized format initially, with details to

follow. Only meaningful information should be included (Weitl, Su9I, Kammerl, &

Freitag, 2002).

Nine considerations for designing informational web content were listed by

Weston, Gandell, McAlpine, and Finkelstein (1999), including computer literacy,

computer access, institutional and infrastructure interactivity, navigation, currency

and validity of information, the technological considerations of loading speed, and

bandwidth. Table 2-1 lists various web site elements that have been

recommended for web sites to optimize information acquisition and retention.


Table 2-1. Support in the literature for web site content elements


Web Site Content Elements

Currency of content



Date of last update indicated


New information indicated
Purpose/mission statement

Table of contents
Color pictures


External Validity

Govers et al. (2000), Jung & Butler (2000),
Motley (2000), Innovations in Distance
Education. (1998), Pan & Fesenmaier (2000),
Raward (2001)
Jung & Butler (2000), Kormier (1999), Motley
(2000), Pan & Fesenmaier (2000), Raward
(2001)
Motley (2000), Raward (2001)
Gretzel et al. (2000), Jung & Butler (2000),
Kovacs & Rowell (2001), Raward (2001)
Jung & Butler (2000), Raward (2001)
Gibson (2001), Schuman (1996)









Table 2-1. Continued.


Web Site Content Elements

Navigation tools on all pages



Site map
Main message of the page
presented "before the fold"
Text equivalent for non-text
elements
Information conveyed with
color available without color
Minimum of one link per page

Backgrounds
Navigation bar
Tables
Animated GIFs
Text effects
Clip art
Graphics
Links to further (internal)
resources
Reliability of links to external
resources
Frequently Asked Questions
(FAQ) included

Internal search tool

Affiliate program

Security


External Validity


Govers et al. (2000), Gretzel et al. (2000),
Jung & Butler (2000), Motley (2000), Poling
(1999), Raward (2001), Schuman (1996),
Tierney (2000)
Motley (2000), Raward (2001)
Schuman (1996)

Raward (2001), Schuman (1996)

Raward (2001)

Fesenmaier (2000), Poling (1999), Pan &
Raward (2001), Schuman (1996)
Jung & Butler (2000), Schuman (1996)
Poling (1999)
Kovacs & Rowell (2001)
Gibson (2001), Poling (1999), Schuman (1996)
Innovations in Distance Education (1998)
Schuman (1996)
Poling (1999)
Tierney (2000), Innovations in Distance
Education. (1998), Raward (2001)
Govers et al. (2000), Innovations in Distance
Education. (1998), Raward (2001)
Jung & Butler (2000), Govers et al. (2000),
Kormier (1999), Innovations in Distance
Education (1998), Raward (2001)
Govers et al. (2000), Gretzel et al. (2000), Pan &
Fesenmaier (2000)
Jung & Butler (2000), Gretzel et al. (2000),
Poling (1999)
Jung & Butler (2000), Poling (1999), Pan &
Fesenmaier (2000)









Table 2-1. Continued.


Web Site Content Elements

Can user comments be
recorded?
Is it possible to get feedback?

Is it possible to ask questions?

Contact information available
Help available

E-mail



Bulletin board


Chat room



Guest book / visitor comments

User forum (threaded)

Forms
Fax
Java
Listserv
Flash presentations
Headings stand out on the
page
Reading level appropriate


Page scan-ability


External Validity


Raward (2001)

Gibson (2001), Innovations in Distance
Education (1998), Raward (2001)
Motley (2000), Innovations in Distance
Education (1998), Raward (2001)
Motley (2000), Raward (2001)
Innovations in Distance Education (1998),
Raward (2001)
Kormier (1999), Gibson (2001), Motley (2000),
Gretzel et al. (2000), Innovations in Distance
Education (1998), Pan & Fesenmaier (2000),
Raward (2001)
Gretzel et al. (2000), Innovations in Distance
Education (1998), Pan & Fesenmaier (2000),
Raward (2001)
Gretzel et al. (2000), MacDonald & Caverly
(2001), MacDonald & Caverly (2001),
Innovations in Distance Education (1998), Pan &
Fesenmaier (2000)
Pan & Fesenmaier (2000), MacDonald &
Caverly (2001)
MacDonald & Caverly (2001), Innovations in
Distance Education (1998)
Pan & Fesenmaier (2000)
Innovations in Distance Education (1998)
Poling (1999), Pan & Fesenmaier (2000)
Innovations in Distance Education (1998)
Pan & Fesenmaier (2000)
Raward (2001)

Innovations in Distance Education (1998),
Raward (2001)
Raward (2001)









Table 2-1. Continued.

Web Site Content Elements External Validity

Text simple, concise, clear Govers et al. (2000), Innovations in Distance
Education (1998), Raward (2001)
Visual appeal Jung & Butler (2000), Schuman (1996), Gretzel
et al. (2000)
Format consistency Kovacs & Rowell (2001), Schuman (1996),
Raward (2001)
Information presented in Innovations in Distance Education (1998)
readable blocks
Headings brief and informative Raward (2001)


Technology in Tourism

Tourism destinations are offering web sites that provide travelers with

online travel brochures to assist in planning business or pleasure trips. Govers,

Jansen-Verbeke, and Go (2000) assert that the tourism industry is too focused

on actual booking and sales, when actual booking is only one phase of the

decision to purchase process. Use of the web in information gathering and

planning may be more advanced than for actual booking and payment (Govers,

Jansen-Verbeke, & Go, 2000). The availability of current, up-to-the-minute

information is a primary advantage of travel planning through the Internet

(Schley, 1997), as is the use of multimedia.

Multimedia is an umbrella term that refers to using more than one form of

media in a single media product (e.g., print, sound, animation). Multimedia

encompasses a wide range of technologies, and is considered the technological

wave of the future (Sigala, Airey, Jones, & Lockwood, 2000). Interactive media is

especially suited to the information-intensive travel and tourism industry (Gretzel,









Yuan, & Fesenmaier, 2000), where content delivery must relate to the knowledge

base of the user. Building a bridge between presented material and the visitor's

existing knowledge base is a key factor in successfully using multimedia. As a

rule, a person will learn more if many links are established by the instructional

content to prior experience and knowledge. The establishment of these links is

known as information mapping (McFarland, 1995).

Information Search

The World Wide Web is an important tool for information search and

retrieval (Fui-Hoon Nah, 2004). When making a purchase decision, the tourist

may consult many sources of information. This process is known as information

search. Consumers search for information more so when product costs are high

(Dana, 1994). The types of information sources available are major factors in a

person's decision to purchase a specific tourism product or service. Tourism

products and services have inherent uncertainty associated with them. That is,

tourism products and services are usually expensive and cannot be previewed or

sampled prior to purchase. Therefore, the tourist who travels to a destination for

the first-time commonly conducts information search using several information

sources, including both marketer-dominated promotional materials and objective

consumer information services (Eby, Molnar, & Cai, 1999). Most purchase

decision-making models feature pre-purchase information search as a key factor

(Bieger & Laesser, 2004). A good understanding of information search behavior

is key to planning and decision-making for businesses (Moorthy, Ratchford, &

Talukdar, 1997). There is a cost associated with each consumer alternative, and









that cost may be either time spent or cognitive effort (Meyer, 1982). Cognitive

costs, such as the evaluation of information and consideration of alternatives are

also integral to information search, although the consumer's perceptions of these

costs are subjective and may vary depending on the searcher's mood,

involvement, or education. For this reason, the total cost of information search to

consumers is incalculable. Information search activity inevitably consumes the

searcher's time and energy, in addition to the financial costs associated with

online access (Hodkinson & Kiel, 2003). As a result, we should expect

information search to be utilized only by those people for whom the value of

information search exceeds the cost (Balakrishnan, 1991).

Vogt and Fesenmaier (1998) considered information search behaviors in

the travel and tourism industry, and expanded the role of tourism information

search from a primarily marketing purview into a broader communication context.

Consumers learn from their own experiences, and through visual, verbal, and

sensory stimuli that educate them about products, including tourism destinations.

This information is then stored in memory and used for functional reasons such

as planning or taking trips (see Figure 1-1).

Major factors identified in the tourism literature which influence information

search include price, the composition of the vacation group, the presence of

family and friends at the destination, prior visits to the destination, and the degree

of novelty associated with the destination (Snepenger, Meged, Snelling & Worral,

1990). The distinguishing feature of consumer information search behavior is

that of consumers' prior knowledge of the market and any implications of this









knowledge for the search process. Advertising can influence a consumer's brand

perception and method of information search (Baloglou, & McCleary, 1999;

Gallarza, Saura, & Garcia, 2002; Moorthy, Ratchford, & Talukdar, 1997).

Wilkie and Dickinson (1985) state that an important research goal is a

fuller understanding of the benefits consumers achieve by information search

behaviors, along with more in-depth details of the search process itself. In a

study centered on the multiple information needs of tourism information seekers,

Vogt and Fesenmaier (1998) found that information is primarily collected and

used for functional reasons, such as planning or taking trips. Other, secondary

needs surfaced in the forms of hedonic and aesthetic needs that play a critical

role in the information search process. Hedonic needs allude to the view of

consumers as pleasure seekers and the need for information search to produce

some pleasurable effect on the searcher, such as enjoyment, amusement or

sensory stimulation. Aesthetic needs can be satisfied from both verbal and

visual forms that stimulate visual thinking and imagery, along with fantasizing

about places that seem real and obtainable. They concluded that, to guide

potential tourists toward a purchase decision, tourist information should appeal to

a variety of search needs to first capture the attention of potential travelers.

The model of consumer purchase behavior adapted by Vogt and

Fesenmaier (1998) depict the role of active and passive external sources in

information acquisition and processing. A key external source for travel and

tourism information acquisition is a destination web site. Current technologies

have enabled web sites to act as both passive and active sources of information.









Web site visitor interest can be engaged by utilization of content and support

elements that act as passive information presentation strategies.

Information search behavior is an important concept in the study of

consumer behavior, and the web provides a major search resource for

consumers (Hodkinson & Kiel, 2003). In a study that profiled people who used

the Internet to gather travel information or purchase travel arrangements, Weber

and Roehl (1999) found that travel information searchers were 26 to 55 years of

age, had higher incomes, were in professional or computer-related occupations,

and tended to have more online experience. In terms of time spent, the Internet

was gaining rapidly on more traditional types of media like television and

newspapers.

Understanding the process by which consumers acquire information is

critical for management and marketing decisions (Bieger & Laesser, 2004). The

motivation for modeling how consumers gather information is to better

understand the cognitive processes governing consumer behavior in the

marketplace (Meyer, 1982).

The goal of information search is usually to obtain a specific bit of targeted

information and recall of the full content of information contained in the searched

document is generally not the focus of the reader. As a result, the cognitive

processes needed for limited information retrieval are utilized to a greater extent

than those needed for prose comprehension (Guthrie, 1988).











Incorporating Instructional Message Design into the Consumer Information

Search and Processing Model

Digital technology and interactive multimedia have had a significant impact

in the field of education, where web-based learning environments utilize the

power of the Internet to create a context where learning is both supported and

enabled. The Internet offers many unique features for learners and can provide

a wealth of information that is not readily available elsewhere. Furthermore, the

information available on the web is usually current, presented in meaningful

contexts, and linked to other available information that explores topics more

widely. In addition, the Internet affords opportunities to be more interactive and

collaborative, creating a true global community where technological tools such as

email, listservs, newsgroups, and videoconferencing enable exchange of

knowledge, ideas, and perspectives.

Development of the Internet and high-speed communication systems has

led to new opportunities for delivering information (Teh, 1999). In the field of

education, digital technology can provide a means to utilize teaching methods

that have been demonstrated to be effective, providing ways to engage students

in active learning and offer easier access to vast amounts of information.

Evidence has revealed that when people are actively involved in a self-driven

learning project, they learn more and remember it longer than when they are

passively sitting and listening (Newman & Scurry, 2001). The challenge for web









designers is not only to generate traffic to their web sites, but also to keep visitors

involved using proven design practices (Jung & Butler, 2000).

In the digital environment, technology has already begun to alter the way

people learn in every setting from skill-training sessions to elementary, high

school and college courses, as well as in traditional classrooms (Newman &

Scurry, 2001). Learning behavior theorists believe that adult learning attitudes

and behaviors are built upon childhood learning experiences. Consumers who

are obtaining information from the web do so by exhibiting previously held

learning behaviors in a new and technically mediated environment. The process

of information search, evaluation and decision-making thus involves learning and

learning styles (Hodkinson & Kiel, 2003). Therefore, benefits from learning

strategies that have been proven effective in online classrooms may transfer to

instructional web site design. In either venue, technology will be used to

enhance learning, encourage communication, and provide sources of further

information (Newman & Scurry, 2001).

Effective teaching and e-learning strategies have been widely reported. A

summation of some effective distance learning strategies was published in

Innovations in Distance Education (1998). This included the following suggested

strategies:

1. A comprehensive 24-hour a day, 7 day a week system of support in place.
2. Regular feedback mechanisms maintained.
3. Frequent and meaningful interactions among learners, between learners
and instructional materials, and between learners and the instructor.
4. E-communications technologies should be used as a tool for creating and
maintaining learning communities.









5. Programs should employ creative solutions to achieve the traditional
outcomes desired.
6. Social interactions between and among e-learners enrich the learning
community and should be supported in the design and delivery of the
program.
7. The program should incorporate a technology base that is appropriate for
the widest range of e-learners within the target audience.
8. The selection of instructional media and tools should reflect a thorough
understanding of the added value of the technology.
9. The design should take into account the diversity of potential users.

In the virtual classroom, the emphasis is on the learner or on how the

learner is going to assimilate the material. People have various learning styles,

and that fact must be taken into account. For online learning to be successful

different learning styles must be accommodated (Weitl et al., 2002). An

interactive environment should also be fostered (Gibson, 2001). Important

factors for successful marketing on the Internet include attracting users,

engaging interest and participation, retaining users, learning about user

preferences, and providing feedback (Gretzel, Yuan, & Fesenmaier, 2000).

Metros (2001) found that online learner engagement ranges from passive

interest to dynamic interaction to flow. The term flow refers to states of deep,

intense involvement in challenging activities that do not overwhelm the skills of

the learner (Csikszentmihalyi, 1975). Web site browsing, for example, would

demonstrate passive interest by the site visitor. The tools of interactivity could

then be used to stimulate dynamic interaction where a flow state may eventually

be achieved. Use of interactive web site features such as threaded discussions,

chat, and live web cams can act as active external information sources, and at

the same time provide the dynamic interaction necessary to engage the site









visitor. The information content should be educational and should be written to

inform people about attractions, destinations, and services (Eby, Molnar, & Cai,

1999).

Cho and Fesenmaier (2000) use the term "telepresence" to describe much

the same process. Telepresence is akin to Csikszentmihalyi's (1975) concept of

flow because both states involve a high degree of involvement in a task or

activity to the exclusion of outside stimuli or the awareness of self or time.

Internet Web Site Evaluation

Evaluation of web site effectiveness is necessary, in part, due to the

substantial costs associated with establishment and maintenance of web sites

(Tierney, 2000). In general, the web lacks standardized measures for evaluation,

as the complexity of the web makes standardization a challenge (Gretzel, Yuan,

& Fesenmaier, 2000).

Tabulating direct sales from a site is one method of evaluation, but many

DMO web sites do not sell directly to the consumer and cannot be measured in

this manner. Further, it is well established that visitors to a site are often there

primarily to gather information. It is therefore critical that the information that a

consumer seeks be presented in a clear and memorable fashion if later sales are

to be realized (Tierney, 2000).

A common method of measuring usage of web sites is number of hits,

followed by number of user sessions, and online feedback from site users.

Tracking hits as a measure of web site effectiveness can be misleading,









however, as the amount of traffic to a site may bear little relation to the

effectiveness of the site (Tierney, 2000).

The nature of web sites allows designers to easily and effectively log site

usage, number of hits, and patterns of site navigation. This information,

however, reveals little about visitor characteristics, motives, preferences,

satisfaction, or likely actions that the visitor will take as a result of visiting the web

site. Incorporating interactivity into the web site, however, can enable the visitor

to impart a great deal of personal information back to the host (Tierney, 2000).

Jung and Butler (2000) listed some common measurement methods that

determine the success of a web site, including number of hits, time spent,

booking rate, interactivity, repeat visits, feedback, currency and gateway

partnerships (commercial banners).

A more reliable evaluation statistic is the "user session," defined as a

session of activity (all hits) for a particular user at a particular site. User session

information can be generated by software packages and provide data on visitors'

domain names, region or country of request, organization type, advertisements

viewed, and time spent in various areas of the site. However, no information on

visitor demographics, motivations or likely actions taken from viewing the site can

be known from user session type data (Tierney, 2000).

An evaluation method growing in popularity is the online survey, a

interactive method that can provide detailed data on the web site visitor. A

serious concern about Internet surveys, however, is nonresponse bias. A survey









can only measure data from those who choose to participate, and Internet

surveys can have abysmal response rates (Tierney, 2000).

Another site effectiveness or tracking strategy employed by commercial

web sites is the use of "cookies," which are data files used by web servers to

identify web users and track user's browsing habits (CIAC, 1998). Unfortunately,

cookies cannot identify individual users, and many web surfers simply turn off the

"accept cookies" option in their web browsers to avoid tracking (Tierney, 2000).

According to Motley (2000), one of the most crucial characteristics of a

good web site is "freshness," the driver of web site "stickiness." To retain

customer loyalty, Internet-based companies and web sites seek stickiness-the

ability of a site to attract repeat customers. People expect to interact with a web

site to experience new things or learn current information. Offering some type of

interaction ensures that there is a reason to return to a web site.

Similar descriptions of types of web site designs were offered by business

(Poling, 1999) and educational (MacDonald & Caverly, 2001) researchers (Table

2-2). For education, the primary concern is to select the most appropriate

technology for the teaching and learning situation. Pedagogy comes first;

technology is only utilized to assist the teaching (MacDonald & Caverly, 2001).

Given the mission of the CVB as both a marketing and educational entity,

elements from both business and education should be considered.

Instructional Message Design

Tufte (1990) noted that the principles of information design are universal

and are nonaligned with any particular language or culture. The specific focus of









Table 2-2. Web site evolution by generate

Poling (1999)


Generation
Simple online presence, an electronic
business card
Generation
Polished, professional look, quick and
easy navigation, and clear goals for
results
Generation
Cutting edge marketing techniques such
as affiliate programs, interactivity, and
using animation and Java programming
to enhance functionality and appeal


ion


MacDonald & Caverly (2001)
1 Web Site
Provides online information with links
to further sites
2 Web Site
Incorporates G1 information, but adds
interactivity through email and web-
based forums
3 Web Site
Includes features of G1 and G2, but
introduces synchronous interactivity
through audio/video conferencing and
chat rooms


instructional message design is linking learning theory and educational practice

(Fleming & Levie, 1993). The basic tenets for linking educational research and

practice are instructional message design principles that establish links between

instructional conditions and learning outcomes.

There is a difference, however, between descriptive and prescriptive

principles. Both involve instructional conditions and outcomes. Prescriptive

principles, however, expressly describe learning processes, whereas, descriptive

principles describe methods of instruction. Prescriptive principles are clearly a

greater value to instructional developers, because they link instructional methods

with learning outcomes. Positive outcomes for message design, then, are

expressly enabled by principles that illuminate those instructional methods that

are most likely to result in the instructional designers intended outcomes

(Fleming & Levie, 1993). The essential instructional message design question is









this: what conditions should be created that are most likely to result in the desired

outcome (Fleming & Levie, 1993)?

Reigeluth (1983) defines instructional design as the process of deciding

what instructional methods are best for enabling intended changes in knowledge

or skills. A "message" is a display of text and graphics or other signs produced

expressly to modify the psychomotor, cognitive, or affective behavior of people.

"Design" refers to the process of analysis and synthesis that starts with an

instructional objective and finishes with a blueprint for that objective's realization.

Message design principles can be best utilized as guidelines for the

display of a visual message (Pettersson, 1999). Winne (1995) suggested that

learning is affected by the presentation of information, and that learning improves

when the same information is delivered in different formats (words, pictures,

sounds). The study of instructional message design allows designers to make

useful choices from a myriad of instructional methods and materials to promote

information acquisition and retention (Fleming & Levie, 1993).

Complicated language, texts, pictures, or graphic design will hinder the

learner in understanding the intended message. Informational text and graphics

should be designed so that they are easy to read. The goal of information design

is clarity of communication. For learning to occur, however, the learner must be

interested and curious about the subject matter (Pettersson, 1999). Learning

increases when learners' attitudes are more positive (Simonson, 1984).

Message designers must ensure that the intended message is not skewed

by the packaging of the information. Many multimedia packages deliver more









information than users can absorb at one time, thus hindering the educational

process. To minimize learner frustration, the quantity of information presented

should be limited to digestible bites (McFarland, 1995). Impositions on working

memory from nonlearning events can also be reduced to the effective use of

different media. Still another strategy to reduce the imposition of short-term

memory caused by complex message comprehension is to make instructional

messages simple and easy to understand (Farquhar & Surry, 1995).

For delivery of content in online learning, the structure and presentation of

content should be similar to the structure and presentation of hypermedia on the

Internet. According to Weitl, Su9I, Kammerl, and Freitag (2002), there are two

major problems inherent in using the web for learning: (a) loss of overview as a

consequence of limited screen area, and (b) the shortened attention span of the

learner to fatigue from reading online. Loss of overview occurs when the volume

of information that can be presented on-screen is limited compared to print

media. To mitigate loss of overview, web content should be presented in a

generalized format initially, with details to follow. Only meaningful information

should be included.

Understanding of instructional message design is incomplete without

understanding the role of motivational appeal. The ultimate success of an

instructional message will depend upon both the learner's willingness and ability

to receive and understand the message (Fleming & Levie, 1993). In the case of

CVB web sites, designers must consider how the destination message can be

delivered in such a way that will motivate the user to seek additional information.









Goals of Effective Message Design

Effective, supportive message design should enhance information

acquisition and retention by reducing cognitive load, or what Miller (1956) would

describe as the restraint on the amount of information we are able to process and

remember. Design elements can enhance or inhibit motivation and attention.

Sound message design always serves to maximize motivation and attention,

while eliminating elements that detract from these goals.

Motivation. The primary goal of instructional message design is to utilize

principles of motivation that result in increased motivation to learn. Gagne (1985)

lists motivation as the first event of instruction, but instructional designers should

consider motivational appeal in every part of the message. Motivation can be

defined as that characteristic which determines the magnitude and direction of

behavior. In this definition, magnitude is the degree of effort, while direction can

be defined as optimum orientation toward goal (Keller & Burkman, 1993).

Motivation in learning is dependent on the learner's personality and

learner's perception of value and the difficulty of learning objective. Instructional

message designers confront this reality by accepting, in some part, responsibility

for motivation in learning. While it is true that an instructor does not totally

control a learner's motivations, poor instruction can demotivate just as great

instruction can inspire learners (Keller & Burkman, 1993). Attitudes are also

related to learning and can be changed by technologically mediated instruction

(Simonson, 1984).









For the learner, motivation is a primary determinant in the outcome of

learning events. For this reason, motivation should be considered throughout the

instructional message design process. The specific focus of instructional

message design is linking learning theory and educational practice. The study of

instructional message design allows educators to make useful choices from a

myriad of instructional methods and materials (Fleming & Levie, 1993).

Learners are more likely to feel confident and motivated about

instructional materials that convey a favorable first impression than those that do

not. Most motivational text and graphic principles relate to gaining and

maintaining attention, relating content to learner interests, goals, or past

experience, and building and maintaining learner confidence in the ability to use

the content (Keller & Burkman, 1993). We can assist the learner in

understanding the message by presenting it as an organized set of ideas (Witt,

1981). The organization and delivery of the message affects the perceived

credibility of the message (Simonson, 1984). According to Weitl et al. (2002), to

mitigate loss of overview, web content should be presented in a generalized

format initially, with details to follow. Only meaningful information should be

included. To mitigate the probable short attention spans of the learners, web

content should be presented in small volumes.

Messages that tell a story have a greater persuasive impact than

messages that simply provide information. When information is structured and

presented logically and intelligently, it is more likely to be perceived as

persuasive. One effective technique that facilitates an increase in perceived









credibility is to use actors demographically similar to the target audience

(Simonson 1984). Learner attitudes are changed when mediated messages are

germane and technically well done, rather than irrelevant and poorly produced.

Mediated instruction that is perceived as realistic relevant and technologically

stimulating is more persuasive to the learner (Simonson, 1984).

In many studies concerning use of media and instruction, achievement,

and not motivation, has been the dependent variable. Other media research has

focused on the impact of short messages with limited content, such as TV

commercials and print advertisements, and therefore does not equate easily to

instructional design (Keller & Burkman, 1993). Motivation plays a critical role in

the CVB mission of destination message delivery; a CVB's web site must

motivate the visitor to seek additional information.

Attention. Industrialized countries are also mass media societies. Living

in a mass media society means exposure to a constant stream of information via

media that permeates home, school, and work environments. It is nearly

impossible to avoid this constant stream of information. But, in this sea of

information bytes, it may also be very difficult to obtain the specific information

that we really seek. Because audio, text and visuals constantly compete for our

attention, it is possible to miss the information that we are really interested in.

Our attention is directed not only to those topics that we are interested in, but

also to sounds, to objects that move, to objects that are large, bold, new, or have

the right color, shade or contrast that enable them to stand out from the

environment (Pettersson, 1999).









Urbina (1994) defined attention as readiness on the part of the receiver to

perceive stimuli in the environment. Attention determines which environmental

events we become conscious of. Gagne (1977) averred that gaining the

attention of the learner was the first step in instruction. In similar fashion, the first

problem for an information designer is to gain the attention of individuals.

Informational material must capture the attention of the intended audience. In

addition, the message must be understandable to the intended audience; that is,

the content of the message must in some sense agree with the learner's attitudes

and beliefs and motivate the learner to understand the message. Attention can

be controlled automatically or by specific instruction; an information designer can

select from different guidelines and design principles in order to affect the

reader's attention (Pettersson, 1999).

Attention is the mechanism that determines which stimuli will be engaged

and which stimuli will be ignored (Carlson, 1993). Mayer (1993) stated that for

learning to occur, the learner must first pay attention to the relevant information.

To accomplish this goal, instructional designers must effectively draw the

learner's attention to the relevant information. Attention is never objective; it is

always a selective process (Pettersson, 1999). Pashler (1995) noted that full

attention enables the fast and efficient completion of tasks. Much of attention is

controlled by will, a conscious internal process. External stimuli can also affect

control of attention. External stimuli that are new or intense can sometimes

control our attention for short periods of time. If the same external stimulus is

repeated or becomes continuous, its command of attention will diminish. The









use of effective message design principles can enhance the CVB web site's

capacity to get and maintain the user's attention.

Visuals and graphic images are perceived more rapidly than texts

(Fleming & Levie, 1978). An often-repeated opinion on the function of visual

images concerns getting attention and maintaining attention (Pettersson, 1998).

Among the visual devices with attention-getting capabilities are split screens,

shading and contrast, text and graphics, and zoom lens type movements that

illuminate or emphasize specific details. Although pictures and graphics may not

carry the main weight of a message, they are useful in attracting the direct

attention of the learner (Witt, 1981). The primary function of still pictures is to

attract and hold attention, to facilitate, persuade, illustrate, clarify, motivate or

reinforce (Pettersson, 1998).

Fleming and Levie (1978) stated that an appreciation of the movements of

the human eye is important to message designers understanding of getting and

maintaining attention. We scan the things we look at (Pettersson, 1999). Bevan

(1998) argues that web site visitors rarely read web pages word for word.

Instead, they scan to find the information they need. Message designers can

make texts scanable by using meaningful headings, bulleted lists, and

highlighted key words. Bevan also cautions that unnecessary whitespace may

make scanning of the text more difficult, resulting in an imposition on gaining

attention. Such impositions result in non-optimal learning environments that can

lead to shortened attention spans among learners (Weitl et al., 2002).









It is important for designers of text-based content to design text that

creates a favorable first impression. A favorable first impression enables

potential learner's to scan the content and to feel confident that they can easily

read and understand messages. A favorable first impression or image may help

to gain and keep learner attention and build learner confidence that they can

successfully navigate and use the content (Keller & Burkman, 1993).

It is advisable, therefore to make the initial perception of printed content

easy-to-read and use in order to build confidence and maintain attention.

According to Keller and Burkman (1993), well-organized and explicit text also

maintains learner attention and builds confidence. Additionally, using readable

language, a familiar font and font size, and appropriate color graphics helps gain

and maintain learner attention. To further mitigate probable short attention spans

of learners on the web, content should be presented in small volumes (Weitl et

al., 2002).

Gaining and keeping the attention of the learner is a useful tool for

reducing the imposition of nonlearning events on working memory. Gagne,

Briggs, and Wagner (1998) report that changing stimuli and appealing to learner

interests can gain and maintain attention.

Message Design Elements Enhance Attention and Motivation

The most critical elements of supportive message design are the text and

graphics used to convey the message. Numerous factors contribute to the

effectiveness of text and graphics. It is possible to use text and graphics as

motivational tools and as methods to get and maintain the learner's attention.









The appropriate use of textual features such as font, graphics, and page design

can increase attention and influence motivation to learn, even though these

methods may be secondary to teaching methods and content (Keller & Burkman,

1993).

Text. Text is the most common medium of instruction. Text

comprehension is a constructive process in which the learner integrates the

content with his or her own experience (Pettersson, 1999). When designing text,

it is useful to consider your target learner's reading skills and interest in reading.

Most principles that are successful with below-average learner's also can be

applied to above-average learner's, although the converse is not true (Keller &

Burkman, 1993). Although readability typically refers to the ease with which the

language in text is read and understood (Chall, Bissex, Conrad, & Harris-

Sharpies, 1996), many other aspects of text contribute to its readability.

The phenomenal growth of the World Wide Web, coupled with our

increased use of word processing programs for creating documents, has

increased the volume of text that we read from computer screens (Dyson, 2004).

When reading a computer screen, many learners scan textual materials prior to

reading them. Dyson suggests that Web pages, in particular, are frequently

scanned rather than read in great detail. The pre-reading scan is generally

focused on identifying the topic of the material and assessing its level of interest.

For many typical learners however, this pre-reading scan serves a secondary

function to assess the level of difficulty in reading and using the content. The

outcome of that assessment can determine whether learner's approach the task









of reading optimistically or pessimistically, and may cause some learner's to

abort reading at that point (Keller & Burkman, 1993).

The formatting of text-based content can determine whether it is perceived

as easy or complex. Relatively short blocks of text and concise messages tend

to suggest fairly easy reading to some learners, and that initial easy-to-read

perception builds confidence and maintains attention (Keller & Burkman, 1993).

Large sections of text should be available for printing or downloading, as most

users will not read large amounts of text online (Bevan, 1998).

Readability of text is, in part, dependent upon a balance among font, font

size, leading (i.e., distance between lines of type), and column width. For this

reason, most graphic designers follow traditional typographic rules when

constructing text. As a result, readers have become accustomed to similar

standards in typesetting, a familiarity that has become a reader expectation

(Keller & Burkman, 1993).

Thousands of fonts have been developed and most instructional designers

agree that some fonts make text more readable than others (White, 1987).

Common fonts are easier to read than uncommon fonts (Benson, 1985; Tinker,

1963), and, for print, serif fonts are easier to read than san-serif fonts (Tinker).

Changes in fonts or style can indicate changes in importance or purpose. Font

selection and typographical techniques can focus the reader's attention on main

ideas, key passages or important concepts. Multiple type styles can be visually

confusing (Pettersson, 1999).









To make text easier to read, maintain user attention and increase

confidence, web designers should choose a font that is both widely used

(familiar) and legible. Additionally, the selection of a font that is widely used

increases the chances that the font will be installed and commonly available on

most computers (Keller & Burkman, 1993). Designers can create the perception

of easy reading by dividing lengthy blocks of text into visually distinct segments

(Keller & Burkman; Smith & Orr, 1985; White 1987). Short sections of text are

preferable (Bevan, 1998). As a rule, wider columns of text should have larger

font size and increased leading (Keller & Burkman).

Running text should be in lowercase letters, as text that is printed in all-

caps reduces reading speed (Henny, 1981; Poulton & Brown, 1968). Lines of

text should be composed of 8 to 10 words in 12-point font to make text easier to

read, maintain reader attention and increase confidence. Lines of text should

also be presented in a fairly open, rather than constrained display (Keller &

Burkman, 1993). Pettersson (1991) notes that black print on white paper is over

10% more efficient in learning than white text on black paper. The most legible

combination of text color and background is black text on a light yellow

background.

It is widely accepted that the perception of text-based content is enhanced

by good structure and organization. When presenting text-based content, the

terms structuring and organizing refer to the logical sequencing of topics and

subtopics. Text organizers such as headings and the use of whitespace are

devices that content providers use to make the structure of text more explicit









(Keller & Burkman, 1993). The terms structuring and organizing also refer to

relationships among content elements, such as illustrations and blocks of text,

that are readily apparent, so that content elements that are not related are not

mistakenly perceived to be related (Keller & Burkman, 1993).

Well-organized and explicit text maintains the reader's attention and builds

confidence (Keller & Burkman, 1993). It is evident that explicit text organizers

also make a text appear to be easier to comprehend, and this can be useful as a

motivating tool in and of itself. Graybinger (1985) found that the more structured

and organized text based content appears, the more students seem to prefer it. It

is also helpful to illustrate the meaning of text with supportive pictures or other

graphics (Witt, 1981).

External text structure refers to the devices that focus the reader's

attention on specific parts of the text (Jonassen & Kirschener, 1982). For

example, headings should attract attention, be relevant and identify subject

matter. Headings indicate the relative importance of items within the text.

Typographic queuing refers to the use of bold, italic or underlining to draw

attention to key items in the text (Pettersson, 1999). According to Cisotto and

Boscolo (1995), paragraph headings improve learning, but underlining of key

information does not.

People who write newspapers, magazines and other mass distributed

media generally use a writing style that incorporates the active voice and short,

familiar, image provoking words that are targeted to typical readers (Keller &

Burkman, 1993). Message designers should keep text messages simple and









closely tied to the supportive graphics (Witt, 1981), and use a readable writing

style to maintain reader attention and increase confidence (Keller & Burkman,

1993).

Graphics. The United States could be described as a visually illiterate

society. Visuals are used for marketing splashes and sensationalism and for

entertainment. Recall of pictures is better than recall of text (Gagne & Rower,

1969) and visuals and graphic images are also perceived more rapidly than text

(Fleming & Levie, 1978). Although visuals are generally easily interpreted, they

can seldom convey the same detailed information of text. Where visuals are

vital, visuals should adhere to their role to support and augment textual content

(McFarland, 1995).

Lester (1995) suggested that the reason visuals are a powerful form of

communication is that they elicit both intellectual and visceral responses. In

traditional paper publication of text materials, visuals were sometimes sparingly

used due to the high cost of producing color pictures. The cost of reproducing

color pictures, however, is not a factor with materials that are distributed

electronically over the Internet. This advance in technology played key role in

allowing visual information to be distributed worldwide in a cost-effective manner.

To achieve the full educational benefits made possible by using pictorial

information in teaching and learning, it is necessary for graphic presentations to

be done with an understanding of how people learn from graphics (Goldfarb &

Kondratova, 2003).









Text must support graphics and vice versa. When both text and graphics

convey exactly the same information, users may become bored or distracted.

When both text and visuals are employed, designers should ensure that the

intended messages are not redundant. Text and graphics should be

complementary, offering slightly different views of related information to promote

learning (McFarland, 1995). Further, the spatial context of sound instructional

design ensures that learning is enhanced when text is placed nearer, rather than

farther, to corresponding pictures and graphics (Mayer, 1993).

Illustrations should be used to augment and complement textual content.

A lone picture may be inappropriate for presenting detailed information

(McFarland, 1995). Text containing relevant illustrations is more conducive to

learning than text alone. Accompanying graphics are particularly adept at

illuminating spatial relationships described in the text (Goldfarb & Kondratova,

2003).

Fleming and Levie (1978) noted that in graphics containing multiple

figures of uniform color or shade, a figure of a different color or shade will attract

attention. Even if color does not add important information to a graphic image, it

may still facilitate information acquisition and processing by causing the user to

pay more attention or have more interest in the image (Pettersson, 1998).

Messages must be readable at an appropriate distance, or when seen for

short periods of time. For example, a driver in a rapidly moving vehicle may only

have a short time to read a sign. Therefore, the message on the sign must be

processed and understood quickly. The meaning of the sign should be









immediately obvious. Where appropriate, color combined with shade and

location can be used to gain attention. Color may also be used to imply meaning.

Red, for example, is commonly used in warnings (Pettersson, 1998). Hartley

(1987) observed that colorblindness is more prevalent among men than among

women, with as many as 8-10% of the male population affected, versus only 1%

of the female population. Red-green colorblindness is the most common.

Unfortunately red and green are also used very often as key colors in symbols

and graphics (Pettersson).

Regular and unfamiliar shapes tend to garner more attention than basic or

familiar shapes. Most people can readily perceive basic shapes. Shape

constancy is our shared tendency to perceive shapes as the same despite

changes in distance, angle, and lighting. This constancy and perception is one

reason that basic shapes are common in the design of symbols and icons

(Pettersson, 1998). Evans, Watson, and Willows (1987) pointed out that

graphics can change from being engaging motivators to engaging distracters.

Pictures may be ignored when an overload of pictures is provided in support of a

single message. Pettersson explains that when too many graphics are used,

readers tend to ignore some of them, producing an effect that is the opposite of

attention.

The type of picture that is displayed is important in comprehension

(Pettersson, 1999). Tufte (1983) stated that graphical competence requires three

different skills: substantive, statistical, and artistic. In spite of this, most graphic

work is under the sole control of artistic personnel (Pettersson, 1998).









Designers should always ask whether or not the illustration or visual

matches the text. Pettersson (1998) found clear differences between the

intended and perceived messages. One way to minimize this difference is to

caption pictures to support their intended interpretations.

Graphics and illustrations should be carefully chosen to augment the

understanding of the text. If illustrations or graphics are unrelated to the

accompanying text message, readers may become confused (McFarland, 1995).

Although pictures and graphics may not carry the main weight of the message,

they may attract the primary attention of user. Designers should keep text

messages simple and closely tied to supportive graphics (Witt, 1981).

Cognitive Load Theory

Cognitive load theory as described by Sweller and colleagues (Sweller,

1994; Sweller, Van Merrienboer, & Paas, 1998) reiterates the multiple

components description of working memory. In addition, a distinction is made

between intrinsic and extrinsic cognitive load, where intrinsic cognitive load refers

to the innate difficulty of processing the information itself and extrinsic cognitive

load refers to the format or methods used in presentation of the information.

Intrinsic cognitive load, then, is dependent on prior knowledge and experience,

among other factors, whereas extrinsic cognitive load can be affected by

instructional message design. The value of instructional message design is its

ability to decrease extraneous cognitive load and direct working memory

resources to the learning process itself by limiting the effect of the presentation

features of instructional material (Tabbers, 1999).









According to cognitive load theory, the limitation of working memory

capacity is a key issue in the learning process. For instructional designers,

content must be presented in such a way that the learner's working memory is

focused on relevant information and is not taxed by irrelevant or distracting

features of instructional material. Where both verbal and visuals are presented,

the learner must mentally integrate text and pictures simultaneously, resulting in

a high cognitive load (Tabbers, 1999). Physical integration of visual and verbal

content can decrease cognitive load and increase learning (Chandler & Sweller,

1991).

A learning event occurs when a learner is consciously interacting with

subject matter. Learning events include mentally organizing, categorizing,

visualizing, and repeating information to facilitate its encoding into long-term

memory (Farquhar & Surry, 1995). The process of mentally organizing,

categorizing, visualizing, and repeating information is simplified and aided by

instructional message design.

Nonlearning events are those interactions that are not beneficial to the

learning process. Any distraction from the instructional message that causes the

learner's attention to stray can be said to be a nonlearning event. Nonlearning

events in an education model can be said to be similar to noise in the

communication model. Nonlearning events inhibit the processing of the

instructional message into long-term memory by occupying the limited capacity of

short-term memory (Farquhar & Surry, 1995). Sound instructional message

design minimizes the occurrence of nonlearning events.









Message comprehension events occur when learners use their senses to

receive information from an audio, video, or text based instructional message.

We can define a "typical" learner as one in the mid range of motivation, neither

over or under motivated (Keller & Burkman, 1993). Message comprehension

events compete with nonlearning events as the focus of short-term memory. As

the short-term memory requirements for message comprehension increase,

processing into long-term memory decreases. Effort required to comprehend a

complex message, or attention distracted by nonlearning events, will impede

processing of the instructional message. That is, the combination of nonlearning

events and message comprehension events impede learning events (Farquhar &

Surry, 1995).

Central to many models of cognitive processes are two distinct memory

states, working or short-term memory, and long-term memory. Working memory

is roughly analogous to awareness and is the memory state containing the

information that is currently in process. Long-term memory, on the other hand, is

roughly analogous to a computer hard drive holding a great deal of information

that is stored in an organized fashion (Farquhar & Surry, 1995). To be retained

in long-term memory, information must first be processed by working memory.

However, working memory is limited in both its capacity and duration, which can

hinder the learning process (Cooper, 1998). Unlike working memory, long-term

memory holds information for extended periods of time (Farquhar & Surry, 1995).

According to Mayer (1992), learning occurs with the selection and

organization of relevant content, and the integration of the new content with prior









knowledge. Working memory plays a key, albeit limiting, role in the selection,

organization, and integration of the new content. Working memory is the

pathway between long-term memory and the outside world and, as such, is the

place where the instructional message and prior knowledge are integrated

(Tabbers, 1999). Working memory, therefore, is at the center of the learning

process. Because working memory is limited (Miller, 1956), high cognitive load

or strain on working memory will inevitably impede the learning process. Miller

described the limits of working memory as the number of chunks (seven plus or

minus two) of processing capacity that working memory possessed. The

"chunks" however could contain variable amounts of information and could be

influenced by several perceptual dimensions. As a result, the addition of

elements such as color, size, or spatial location can increase the processing

capacity of working memory.

The goal of educational theories relating to cognition and instruction is the

generation of useful instructional techniques (Sweller & Chandler, 1991).

Cognitive load theory describes the information processing system. The essential

tenet of cognitive load theory is that learning outcomes can be improved through

greater understanding of the role of working memory in the learning process

(Cooper, 1998). Better understanding of the data and the usefulness of its

predictive power are key elements of theories in education, but unless these

elements are followed by improved teaching methods and then the ultimate goal

of education theory is missed (Sweller & Chandler, 1991).









In the field of education, instructional message design is used to create

environments where cognitive load is reduced and learning is enabled.

Evaluating web sites according to instructional message design principles is a

method of appraising web site effectiveness, but in the commercial sector such

techniques have not typically been used.

Complex instructional messages can impose on working memory by

presenting multiple stimuli resulting in a sensory overload that impedes

processing from working memory to long-term memory. When impositions on

working memory are reduced, a greater degree of cognitive engagement results.

There are various instructional message design strategies designed to increase

cognitive engagement through reduction on impositions on working memory

(Farquhar & Surry, 1995).

Use of supportive message design elements, such as effective text and

graphics, increases attention and motivation, which, in turn, reduces cognitive

load. Reduced cognitive load enhances the processing of the intended message

into long-term memory. Figure 1-2 illustrates these relationships.

Application of Cognitive Load Theory and

Instructional Message Design Principles to Tourism Information Search

CVBs have, as their central mission, an obligation to present information

about destinations they represent in a manner that facilitates information

processing by the consumer. This warrants consideration of the role of

supportive message design in the reduction of cognitive load and, thus, in the

facilitation of information processing for tourists. Figure 2-1 represents how









supportive message design fits within a consumer information acquisition and

processing model. Vogt and Fesenmaier (1998) characterized information

processing as a critical link in the model. When information is acquired via the

Internet, supportive message design should reduce cognitive load of the user,

thereby facilitating efficient information processing.


Stage 1


Stage 2 Stage 3 Stage 4

Search for Additional Information


Stage 5


Figure 2-1. The role of supportive message design in the facilitation of consumer
information acquisition and processing.


The process of information search and evaluation involves learning and

learning styles (Hodkinson & Kiel, 2003). The primary responsibility of an

informational web site is to meet the expectations and needs of visitors by

creating an environment where learning can occur (Hubbard, 1998). When used

as an instructional technology, therefore, web pages must be designed to provide

an environment that can enable a successful learning experience.









Instructional web sites can be made more effective through utilization of

the principles offered through the study of instructional message design. The

focus of instructional message design is linking learning theory and educational

practice by prescribing learning conditions that are likely to result in desired

outcomes (Fleming & Levie, 1993).

Effective message design should enhance information acquisition and

processing by reducing cognitive load. Design elements can enhance or inhibit

both motivation and attention. Sound message design serves to maximize

motivation and attention, while eliminating elements that detract from these

goals.

The most critical elements of instructional message design are the text

and graphics elements used to convey effective messages. Both text and

graphics can be used as motivational tools and as methods to get and maintain a

learner's attention. The appropriate use of textual features can increase attention

and influence motivation to learn (Keller & Burkman, 1993).

According to cognitive load theory, the limitation of working memory

capacity is a key issue in the learning process. For web designers, content must

be presented in such a way that the user's working memory is focused on

relevant information and is not taxed by irrelevant or distracting features of

instructional material. The value of instructional message design is its ability to

decrease extraneous cognitive load and direct working memory resources to the

learning process itself by limiting the effect of the presentation features of

instructional material (Tabbers, 1999).






57


By providing information through supportive message design, CVBs can

enhance the effectiveness of their destination message. This study examines

use of supportive message design elements on CVB web sites. Chapter 3

outlines the methods used in the study.















CHAPTER 3
METHODS

The purpose of this study was to investigate the viability of travel and

tourism web sites as instructional environments, and to determine whether

geographic differences exist with respect to web site utilization of instructional

message design strategies on those web sites. That is, this study examined

whether and how effective instructional message design is used by convention

and visitors bureaus to deliver destination message via the Internet. Specifically,

do CVBs utilize the tenets of instructional message design? This chapter

includes a description of sampling procedures and the population of web sites.

Succeeding sections of this chapter include details of the research design and

treatment of the data. The results of this study can be used to evaluate web sites

in the context of information retention and assess the viability of travel and

tourism web sites as instructional environments.

This study was conducted over an eight-week period during the spring and

summer of 2006. The study was designed to answer the following research

questions:

1. What is the prevalence of message design elements on CVB web sites?

2. Which message design elements are CVB web sites using effectively to
deliver destination messages?

3. How many underlying instructional message design conditions that
enable the desired retention outcomes are present on CVB web sites?









4. Do regional or geographic differences exist in the use of message
design elements on CVB web sites?

5. What web page elements do web design professionals believe to be
most important for destination message delivery?

The following sections provide details about the quantitative and qualitative

research methods that were used to carry out this study.

Methods: Research Questions 1-4

The purpose of Research Questions 1-4 was to examine the use of

message design elements on CVB web sites. The following sections outline the

procedures employed to answer these questions.

Item Review

The initial pool of message design items was gathered from a literature

review of published articles, from both popular and scientific sources, related to

text and graphics in instructional message design. A list of 93 elements was

compiled. A review of the professional literature provided external validity for 70

of the message design elements, and the list was subsequently reduced to that

number. Further review of the human-computer interaction literature provided

external validation from studies (similar to message design) conducted using

computer screen (i.e., not print) media. The resultant item list was reviewed by

two design experts: Dr. Lee Mullally (instructional message design) and Dr.

Benjamin Lok (human-computer interaction). The experts provided input to

improve each item's clarity, readability, and appropriateness, and to recommend

removal of items deemed redundant. A final list of 46 variables (Appendix A)

was used in this study.









Sample

CVB web sites were chosen as the unit of analysis due to their role as

providers of information for the tourism industry. A population (N=1,499) of CVB

web sites was identified from a listing of worldwide sites stratified into nine

geographic regions and on the web site TourismBureaus.com

(http://www.tourismbureaus.com/default.asp). The final sample was randomly

selected from each of nine world regions (Table 3-1). Using a power analysis to

determine a minimum N and adding a sufficient number to allow for attrition, a

sample of 625 CVB web sites (from a population of 1,499) was randomly

selected for data collection. Due to the transient nature of web sites, some of the


Table 3-1. Sample from CVB world regions


Number of listed CVBs per
geographic region (total = 1499)

Asia (42)
Africa (20)
Central/South America (30)
Canada (77)
Europe (153)
Caribbean (41)
Middle East (15)
United States (1082)
Oceania/South Pacific (39)
Total


Sample N for
region

30
18
25
55
109
30
13
277
31
588


% of Region Confidence
Sampled interval

71 4.22
90 3.27
83 3.55
71 3.10
71 2.20
73 4.09
87 4.48
26 2.21
80 3.52


links were not functional at the time of data collection. If another link from the

same geographic strata was available, a missing link was replaced with another

randomly selected site. Using this procedure, the resulting final stratified random









sample size was 588. This sample was examined to answer research questions

1 through 4, whereas, research question 5 was answered through a Delphi study.

The final sample of 588 CVB web sites was examined using quantitative content

analysis, a method well suited to the study of communication (Babbie, 2001), to

assess the number and types of instructional message design elements utilized.

Data Collection Procedures

The source of content assessed from each web site was limited to the

initial information page on each web site to standardize data collection. The CVB

web sites were visited using the web browser Mozilla Firefox 1.5 for Windows.

The computers used in this study were set at default preferences for display

properties, using 32-bit color and a monitor resolution of 1024x768 (OneStat,

2005).

Web sites were assayed for content using a checklist-type coding sheet

(Appendix A). During data collection, each element was assessed on a Likert-

type scale ranging from 1 to 4, where 1 = poor, 2 = fair, 3 = good, and 4 =

excellent. If any of the listed items were not applicable to the web page under

scrutiny, the N/A box was checked.

Reliability

The reliability of the scale items was ascertained using SPSS to indicate

the precision of the measurement, or assay how large the errors of measurement

were likely to be. The reliability coefficient, Cronbach's Alpha (a measure of

internal consistency based on the average inter-item correlation), ranges

between 0 and 1, where 1 denotes 100% error-free measurement.











Interrater Reliability

Three raters conducted the evaluation of the web sites. The researcher

served as one of the raters. The two other raters were adults who (a) were

experienced with computer use, (b) had backgrounds in business, and (c)

participated in a two-hour training session conducted by the researcher. A

random sample of 200 sites was assessed by all raters to establish interrater

reliability.

Stemler (2004) avers that the practice of depicting interrater reliability as a

single, unified concept is misleading, and that statistical methods for computing

interrater reliability can be more accurately classified into three categories:

consensus estimates, measurement estimates, and consistency estimates.

Researchers are cautioned that concomitant with different approaches to

estimating interrater reliability are implications for how ratings across multiple

judges should be summarized.

Consensus estimates of interrater reliability, such as percent agreement,

assume that the data reported by reasonable observers should reflect exact

agreement about how to apply a scoring rubric to observed data (Stemler, 2004).

Consequently, if raters can come to exact agreement on how to apply the rating

scale, then the raters may be assumed to share a common interpretation of the

scoring rubric. Percent agreement has some distinct disadvantages, however.

Methods for establishing interrater reliability by using percentage agreement

between raters fail to account for the role of chance in recording data. In content









analysis, some chance is bound to occur, with resultant inflation of agreement

percentages, especially where the number of raters is low (MacLennan, 1993). A

further disadvantage of determining interrater reliability by percentage agreement

is that the amount of time to train raters to come to exact agreement can be

significant, and may be impossible if the number of raters is large. Training raters

to the point of forced consensus may actually reduce the objective independence

of the ratings, and threaten the validity of the study (Linacre, 2002). Finally,

consensus estimates such as percentage agreement can be overly conservative

if two raters exhibit systematic differences in scoring items, but simply cannot be

trained to reach an exact consensus (Stemler, 2004).

Measurement estimates (principle components factor analysis, facet rater

indices and fit statistics) assume that researchers use all of the data available

from all raters, including inconsistent ratings, when computing summary scores,

as it is the accumulation of information, not the ratings themselves, that is

desirable. When using measurement estimates, it is not necessary for raters to

come to a consensus on application of a scoring rubric because differences in

ratings can be estimated in the production of each rater's final score (Stemler,

2004).

Consistency estimates of interrater reliability, such as Pearson's

correlation coefficient, Spearman's rank coefficient and Cronbach's alpha are

based on the assumption that each rater is consistent in classifying the observed

data according the raters interpretation of the scoring rubric. Pearson's

coefficient requires assay of one pair of raters and one item at a time, while









Spearman's coefficients requires all raters to rate all cases; as such, neither

technique was applicable to this data set.

Crocker and Algina (1986) recommended that in studies where multiple

raters are used, a high Cronbach's alpha would produce a consistency estimate

of interrater reliability demonstrating that the bulk of the variance in the total

composite score was true score variance and not error variance. An advantage

of using Cronbach's alpha is its ability to produce a single estimate of interrater

reliability across multiple judges (Stemler, 2004). A Cronbach's alpha score

above 0.7 is acceptable for demonstrating an acceptable estimate of interrater

reliability (Barrett, 2001).

In this study, interrater reliability was established by three methods.

Initially, an analysis of variance on all items using the selection variable raterr"

identified two items that exhibited significant differences between raters. The

variables "56K modem" and "ISDN 128" were removed from further analysis due

to unreliability.

Data recorded on Likert-type items were assayed using principal

components factor analysis. This method can be used as a measurement

estimate of interrater reliability by providing a single statistic, permitting direct

comparison of the raters, and producing an empirical estimate of the extent to

which raters applied the rating scale across variables. If the shared variance is

high (greater than 60%), then this technique indicates that the raters are rating a

common construct (Stemler, 2004). Each rater scored above the threshold level

of 60% (Table 3-2).









Table 3-2. Rater summary scores on primary component

Rater Percent of Variance Explained

1 68.71

2 67.40

3 70.23


Finally, Cronbach's alpha was used to produce a consistency estimate

(Crocker & Algina, 1986) of interrater reliability, supporting the supposition that

the variance in the total composite score was true score variance and not error

variance. The Cronbach's alpha for the Likert-type items was .86, indicating a

high degree of interrater reliability.

Other Variables Measured

In addition to the message design elements addressed by the rating scale,

on each CVB web site evaluated, several other variables were measured, as

well: download time, readability, and legibility. Each of these elements

contributes to the user's experience while visiting a web site.

Download time. Download time for each CVB site was measured by the

Web Page Analyzer v0.961, an online tool provided from Web Site Optimization

(http://www.websiteoptimization.com/services/analyze/). The URL for each site

was entered to determine the download speed for various Internet connections.

Readability. Readability of CVB web sites was determined using an

online measurement tool provided through the web developer resource site

JuicyStudio.com (http://juicystudio.com/services/readability.php#readresults),

where URLs of selected web sites can be submitted for web page assessment by









three reading level algorithms. Although reading level algorithms only provide a

rough guide, the Gunning Fog, Flesch Reading Ease, and Flesch-Kincaid

readability tests can be useful in determining how readable web site content is,

and can give some indication as to whether message designers have developed

content at the reading level of the intended audience.

The Gunning-Fog index determines how many years of schooling would

be required for a reader to understand the web site content, where the lower the

number, the more understandable the content should be (scores over 17 are

reported as 17 to post-graduate level). Results for the Flesch Reading Ease test

are reported on a 100-point scale, where the lower the score, the more difficult

the content is to understand. Content writers are encouraged to develop text that

scores in the 60 to 70 range. The Flesch-Kincaid grade level, like the Gunning-

Fog index, is a measure of the number of years of schooling that would be

required for a reader to understand the text (negative scores are reported as

zero, and scores over 12 are reported as 12).

Legibility. The on-screen legibility or distinctness of the text and

headings was assessed by software provided through the Web Accessibility

Tools Consortium (http://www.wat-c.org). The Color Contrast Analyzer v1.0 uses

a color contrast algorithm to test whether foreground/background color and

brightness combinations are greater than a set range, providing acceptable

screen visibility for both normal vision and three types of vision impairments.

The Color Contrast Analyzer employs formulas suggested by the World

Wide Web Consortium (http://www.w3.org/TR/AERT#color-contrast) to determine









color brightness ((red value X 299) + (green value X 587) + (blue value X 114)) /

1000) and color difference (maximum (red value 1, Red value 2) minimum (red

value 1, red value 2)) + (maximum (green value 1, green value 2) minimum

(green value 1, green value 2)) + (maximum (blue value 1, blue value 2) -

minimum (blue value 1, blue value 2)), where color brightness must be greater

than 125 and color difference must be greater than 500. The Color Contrast

Analyzer also tests color brightness/difference for 3 types of color vision

impairments (i.e., protanopia, deuteranopia, and tritanopia) and indicates

whether they meet threshold requirements.

Treatment of the Data

Data were analyzed using SPSS (Statistical Package for the Social

Sciences) v.12.0 for Windows. Item frequencies, means and percentages were

recorded to compare the descriptive statistics of the sample.

Missing values for Likert-type items were replaced using the SPSS

function Replace Missing Values by selecting the linear trend at point method.

Linear trend at point replaces missing values by regression on an index variable

where missing values are replaced with their predicted values. The linear trend at

point method maintains a degree of overall variance within variables, where

simple mean substitution would not.

Factor analysis, a method of data reduction that attempts to identify

underlying factors that explain the intercorrelations within a set of variables, is

used to identify a limited number of factors that explain most of the variance

observed in a much larger number of variables. Factor analysis of the Likert-type









items was performed to determine whether any sub-scales of the construct exist

within the data collected. The reliability of any sub-scales was individually

assessed. Those subscales with acceptably high reliability (Cronbach's alphas

.7) were selected for inclusion in the final index. The result of this analysis

yielded a multi-dimensional model of a web-based learning environment.

Methods: Research Question 5

To allow for more meaningful interpretation of the results of this study and

to understand its implications for practice, it was important to ascertain which

web page elements web design professionals believe to be most important for

destination message delivery. The continually and rapidly changing nature of the

Internet and of web design practices required interpretation of results via

consultation with practicing professionals, rather than relying on published

research alone. Therefore, research question 5 was addressed through analysis

of a Delphi study conducted on a panel of web design professionals. The Delphi

method is a systematic rigorous and effective method that can produce potent

and valid answers to research questions (Clayton, 1997). The objective of the

Delphi method is to obtain a reliable consensus of opinion of a selected panel of

experts through a series of detailed questionnaires developed through ongoing

feedback.

The Delphi method uses multiple iterations with researcher-controlled

feedback (Murry & Hammons, 1995) from the individual experts and avoids direct

interaction of the individual experts with one another to control unnecessary bias.

The initial questions are intended to elicit the respondents reasoning that went









into their replies, factors considered to be relevant to the study, the respondent's

judgment of these factors, and other information the respondents feel may be

helpful in reevaluating various factors that may provide more illuminating

answers to the primary question (Dalkey & Helmer, 1963). The Delphi method of

controlled interaction among panel members provides an advantage over more

conventional uses of experts, such as roundtable discussions, by encouraging

independent thought and aiding the gradual formation of considered opinion.

Linstone and Turoff (2001) offer a number of situations where the Delphi

technique could be successfully utilized, including situations where

understanding of the topic can benefit from subjective judgments on a collective

basis, and where time and cost make group meetings inefficient.

Computer-based Delphi studies enable respondents to participate in an

asynchronous manner (Turoff et al., 2004). A perceived weakness of the

traditional Delphi technique was the relatively long periods of time required to

administer the method by mail. An e-Delphi system speeds the process of data

collection and feedback and limits data loss from nonparticipation (Chou, 2002).

Delphi Panel Selection

In the Delphi method, group consensus is the goal of the project. The

group is comprised of a panel of experts. According to Scheele (1975) experts

should be selected from among stakeholders (those who may be directly affected

by the topic under study), experts (those who have special experience or

education), and facilitators. The selection of experts (n=4) for this study followed

the following criteria: each expert possessed a special knowledge of web site









creation, including selection of text and graphical content; each expert has been

employed by the tourism industry and charged with web design; and each expert

professed a keen interest in the topic under study.

First Round: Questionnaire

The first round checklist questionnaire (Appendix B) was distributed to

respondents along with detailed guidelines on its completion. The questionnaire

was designed to investigate the importance ratings panel members assign to

various web site elements. The panel was advised as to the depth of responses

sought, and given instruction for rating the items in the study so that all

respondents interpreted the design items in a similar fashion. A four-point Likert-

type response scale was adopted for its familiarity, and to easily identify areas of

agreement and disagreement.

The first round in the Delphi study was designed to illuminate the

respondent's initial positions on the subject at hand. With respect to the topics to

be evaluated, the first round questionnaire identified which items from the

questionnaire the respondents agree upon, and which items provoked

disagreements. A complete record of the first round questionnaire responses

was assembled, and analysis of this record identified agreements and

disagreements among panel members on every item under consideration.

Second Round: Subject Exploration

The second round was predicated on the responses and analysis of the

first round questionnaire. The Delphi process stops when consensus among the

respondents has been reached or when saturation has occurred and no new









information is forthcoming. Subsequent rounds provide the panel with an

opportunity to evaluate both their own responses and group answers from the

previous round.

In the second round, respondents were asked to consider their answers

from the initial round in light of the group responses, and to justify and defend

their positions. In the second round, respondents began to explore why they

believed some web site elements to be more important than others and the

underlying reasons guiding perceptions of relative importance. In the second

round of data analysis, the researcher synthesized new data based on items of

group consensus, and by identifying those items where differences of opinion

existed.

Third Round: Consensus

In the third round, the panel reviewed the remaining areas of dissention

and compared individual views with group views. At this point, most items had

been explored and reevaluated. Analysis of the third round of responses

illuminated some general agreement among the respondents with respect to the

research question given to them, and the reasons for their positions.















CHAPTER 4
RESULTS

Before addressing the research questions, item analysis was performed

following data collection to assess the measurement properties of individual

items. Since a Likert-type scale measures between-individual differences,

recorded values must fall at different levels of the scale or the items would not

indicate any information concerning individual differences on the variable in

question, and thus yield nothing to the overall measurement of the construct.

Each item was analyzed for response variability, response location, and

response discrimination.

Response variability. Penfield's V was used to evaluate variability in

item responses (the index of variation (6) could have been used, but 6 is more

affected than V by items having low response rates). For nearly all variables,

Penfield's Vwas in excess of .90, indicating that the items were doing a good job

of measuring real differences on different facets (Table 4-1). One variable,

"legible font," had medium variability (.75 < V < .90), and could be considered for

revision. Three items, "running text in lower case letters," "avoid jargon/slang,"

and "familiar font" exhibited poor variability and became candidates for removal

from further analysis.

Response location. Items were further assayed for response location

(Table 4-2). Response location indicates whether the majority of individuals lie at









Table 4-1. Response variability of the Likert-type items


Variable N
headings attract attention 588
headings identify subject matter 588
loss text overview 588
whitespace use 588
text distinct segments 588
short blocks text 588
running text lower case 588
lines text 8 to 10 words 588
typographic queuing 588
embedded hotlinks for key terms 588
highlighted key words 588
leading interlinearr spacing) 588
column width 588
avoid jargon, slang 588
same info different formats 588
well organized text 588
explicit text 588
content generalized to specific 588
familiar font 588
legible font 588
font size appropriate 588
multiple type styles 588
captions support graphics 588
captions near to corresponding graphics 588
graphic overview scrolling 588
graphics gain attention 588
favorable first impression 588
picture depiction accurate 588
picture is aesthetically satisfactory 588
picture is technically acceptable 588
picture fits in area 588
picture fits with other pictures 588
picture overload 588
graphics support textual content 588
consistent color scheme 588
realistic colors used in graphics 588
color media preferable to B&W 588


Standard Penfield's Response


Deviation
0.81
0.97
1.10
1.06
.90
0.93
0.46
1.04
1.00
1.20
0.77
0.93
1.01
0.52
1.11
0.96
0.92
1.26
0.40
0.66
0.82
0.84
1.06
1.08
1.10
1.06
1.01
1.00
1.10
0.90
0.99
0.92
1.02
1.05
0.72
0.81
0.89


V
1.08
1.29
1.46
1.41
1.2
1.25
0.61
1.39
1.33
1.60
1.03
1.24
1.35
0.69
1.48
1.28
1.23
1.67
0.53
0.88
1.09
1.11
1.41
1.44
1.47
1.42
1.35
1.33
1.46
1.19
1.32
1.23
1.36
1.40
0.96
1.08
1.19


Variability
Good
Good
Good
Good
Good
Good
Poor
Good
Good
Good
Good
Good
Good
Poor
Good
Good
Good
Good
Poor
Medium
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good









Table 4-2. Response location of the survey items.
Ideal
Variable La
Location
headings attract attention 2.5
headings identify subject matter 2.5
loss text overview 2.5
whitespace use 2.5
text distinct segments 2.5
short blocks text 2.5
running text lower case 2.5
lines text 8 to 10 words 2.5
typographic queuing 2.5
embedded hotlinks for key terms 2.5
highlighted key words 2.5
leading interlinearr spacing) 2.5
column width 2.5
avoid jargon, slang 2.5
same info different formats 2.5
well organized text 2.5
explicit text 2.5
content generalized to specific 2.5
familiar font 2.5
legible font 2.5
font size appropriate 2.5
multiple type styles 2.5
captions support graphics 2.5
captions near to corresponding graphics 2.5
graphic overview scrolling 2.5
graphics gain attention 2.5
favorable first impression 2.5
picture depiction accurate 2.5
picture is aesthetically satisfactory 2.5
picture is technically acceptable 2.5
picture fits in area 2.5
picture fits with other pictures 2.5
picture overload 2.5
graphics support textual content 2.5
consistent color scheme 2.5
realistic colors used in graphics 2.5
color media preferable to B&W 2.5


Absolute
Mean
Difference
2.42 0.08
3.03 0.53
2.79 0.29
3.08 0.58
3.54 1.04
3.42 0.92
3.88 1.38
3.06 0.56
2.23 0.27
2.10 0.40
1.37 1.13
3.46 0.96
3.21 0.71
3.80 1.30
2.20 0.30
3.29 0.79
3.20 0.70
1.99 0.51
3.89 1.39
3.70 1.20
3.20 0.70
3.30 0.80
1.70 0.80
1.71 0.79
3.02 0.52
2.52 0.02
2.55 0.05
3.16 0.66
2.88 0.38
3.32 0.82
3.30 0.80
3.21 0.71
3.38 0.88
2.52 0.02
3.64 1.14
3.63 1.13
3.61 1.11


Response
Location
Good
Medium
Good
Medium
Poor
Medium
Poor
Medium
Good
Good
Poor
Medium
Medium
Poor
Good
Medium
Medium
Medium
Poor
Poor
Medium
Medium
Medium
Medium
Medium
Good
Good
Medium
Good
Medium
Medium
Medium
Medium
Good
Poor
Poor
Poor









the high, medium, or low end of the item scale. It is preferable to have the mean

of the item response distribution lie near the center of the item scale. With a

scale ranging from 1 to 4, it is preferable to have a mean near 2.5. Nine items

exhibited poor response location became candidates for removal from analysis.

Item discrimination. Item discrimination refers to how well the item

distinguishes between high and low levels of the construct being measured. Item

discrimination was ascertained by finding the item-total correlation index. The

item-total correlation measures the correlation between the individual item scores

and the scale score. For a Likert-type scale, an item discrimination score > .5 is

desirable, whereas < .2 is unacceptable. Results of the item discrimination

analysis are presented in Table 4-3. Six of the items have poor item

discrimination, but in no case would the alpha appreciably rise if the item were

deleted. Table 4-4 presents a composite of the item analysis findings and a

determination of which items would be rejected for further analysis.

Thirty-two of the variables had acceptable composite scores for response

variability, location, and item discrimination and appeared to be acceptable

measures of the construct. In some cases (e.g., text in distinct segments), items

had poor ratings on one of the three facets considered in item analysis.

Retention of these items was a judgment call and, because none of the overall

ratings were sufficiently low, the items were not excluded. For five variables,

"running text lower case," "highlighted key words," "avoid jargon, slang," "familiar

font," and "legible font," overall item analysis scores were sufficiently low to

identify those items as poor measures of the construct. These items were









Table 4-3. Item discrimination of Likert-type items (alpha=.86).


Variable
headings attract attention
headings identify subject matter
loss text overview
whitespace use
text distinct segments
short blocks text
running text lower case
lines text 8 to 10 words
typographic queuing
embedded hotlinks for key terms
highlighted key words
leading interlinearr spacing)
column width
avoid jargon, slang
same info different formats
well organized text
explicit text
content generalized to specific
familiar font
legible font
font size appropriate
multiple type styles
captions support graphics
captions near to corresponding graphics
graphic overview scrolling
graphics gain attention
favorable first impression
picture depiction accurate
picture is aesthetically satisfactory
picture is technically acceptable
picture fits in area
picture fits with other pictures
picture overload
graphics support textual content
consistent color scheme
realistic colors used in graphics
color media preferable to B&W


Item-Total
Correlation
.133
.283
.174
.457
.482
.398
.356
.316
.272
.222
.062
.510
.447
.066
.447
.512
.462
.127
.208
.237
.324
.303
.306
.287
.231
.532
.629
.510
.456
.505
.466
.578
.242
.349
.419
.411
.412


Alpha, if
deleted
.863
.860
.863
.856
.856
.857
.859
.859
.860
.862
.864
.855
.856
.863
.856
.855
.856
.865
.861
.861
.859
.859
.860
.860
.862
.854
.852
.855
.856
.855
.856
.853
.861
.859
.857
.857
.857


Item
Discrimination
Poor
Medium
Poor
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Poor
Good
Medium
Poor
Medium
Good
Medium
Poor
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Good
Good
Good
Medium
Good
Medium
Good
Medium
Medium
Medium
Medium
Medium









Table 4-4. Item analysis composite.

Variable
headings attract attention
headings identify subject matter
loss text overview
whitespace use
text distinct segments
short blocks text
running text lower case
lines text 8 to 10 words
typographic queuing
embedded hotlinks for key terms
highlighted key words
leading interlinearr spacing)
column width
avoid jargon, slang
same info different formats
well organized text
explicit text
content generalized to specific
familiar font
legible font
font size appropriate
multiple type styles
captions support graphics
captions near correspond. graphics
graphic overview scrolling
graphics gain attention
favorable first impression
picture depiction accurate
picture is aesthetically satisfactory
picture is technically acceptable
picture fits in area
picture fits with other pictures
picture overload
graphics support textual content
consistent color scheme
realistic colors used in graphics
color media preferable to B&W


Response
Variability
Good
Good
Good
Good
Good
Good
Poor
Good
Good
Good
Good
Good
Good
Poor
Good
Good
Good
Good
Poor
Medium
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good


Response
Location
Good
Medium
Good
Medium
Poor
Medium
Poor
Medium
Good
Good
Poor
Medium
Medium
Poor
Good
Medium
Medium
Medium
Poor
Poor
Medium
Medium
Medium
Medium
Medium
Good
Good
Medium
Good
Medium
Medium
Medium
Medium
Good
Poor
Poor
Poor


Item
Discrimination
Poor
Medium
Poor
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Poor
Good
Medium
Poor
Medium
Good
Medium
Poor
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Good
Good
Good
Medium
Good
Medium
Good
Medium
Medium
Medium
Medium
Medium


Item
Utility
Accept
Accept
Accept
Accept
Accept
Accept
Reject
Accept
Accept
Accept
Reject
Accept
Accept
Reject
Accept
Accept
Accept
Accept
Reject
Reject
Accept
Accept
Accept
Accept
Accept
Accept
Accept
Accept
Accept
Accept
Accept
Accept
Accept
Accept
Accept
Accept
Accept









removed from further scale analysis, leaving a final set of 32 Likert-type items

remaining.

Following the response variability, response location and item

discrimination procedures, items were considered for removal from further

analysis based upon their relative scores. Each of the items that had acceptable

response variability, location and item discrimination were retained, as they

appeared to be good measures of the construct.

Six variables were underrepresented in the sample. Although still useful

for descriptive statistics, these six variables occurred too infrequently on CVB

web sites for meaningful statistical analysis. These six items were: (1)

abbreviations, acronyms and initialisms, (2) metaphors make unfamiliar familiar,

(3) graphics easy to read, (4) meanings of icons immediately obvious, (5)

information w/color also w/out color, and (6) graphics used when messages are

complex.

Results of Data Analysis

Following the item analysis, data were further analyzed to answer each

research question. The following sections present the results of these analyses.

Research Question 1

To answer Research Question 1- What is the prevalence of message

design elements on CVB web sites?-descriptive statistics were compiled to

reflect the appearance of message design elements on the 588 selected CVB

web sites (Table 4-5). Instructional message design elements, in general,

appeared throughout the sampled web sites. A few elements, such as









Table 4-5. Prevalence of instructional message design elements on CVB web
sites.


Message Design Element

white space use
loss text overview
font size appropriate
multiple type styles
familiar font
legible font
consistent color scheme
headings attract attention
favorable first impression
headings identify subject matter
leading/interlinearspacing
column width
well organized text
running text lower case
explicit text
color media preferable to black & white
realistic colors used in graphics
lines text 8 to 10 words
text distinct segments
short blocks text
picture overload
graphic overview scrolling
graphics gain attention
picture is aesthetically satisfactory
picture is technically acceptable
picture fits in area
picture depiction accurate
content generalized to specific
same info different formats
picture fits with other pictures
macrosignals under italics bold
typographic queuing
imbedded hotlinks for key terms
captions support graphics
captions placed near to corresponding graphics


Number of
Sites
585
584
583
583
582
581
581
580
579
578
575
574
574
573
572
572
570
569
567
567
567
565
565
564
564
562
559
555
518
513
498
497
495
487
485


Percent of
Sites
99.5%
99.3%
99.1%
99.1%
99.0%
98.8%
98.8%
98.6%
98.5%
98.3%
97.8%
97.6%
97.6%
97.4%
97.3%
97.3%
96.9%
96.8%
96.4%
96.4%
96.4%
96.1%
96.1%
95.9%
95.9%
95.6%
95.1%
94.4%
88.1%
87.2%
84.7%
84.5%
84.2%
82.8%
82.5%









Table 4-5. Continued.


Message Design Element

graphics support textual content
highlighted key words
avoid jargon, slang
meanings of icons immediately obvious
abbreviations, acronyms and initialisms
graphics easy to read
information w/color also w/out color
graphics used when messages are complex
metaphors make unfamiliar familiar


Number of
Sites
472
471
448
293
244
227
143
138
41


Percent of
Sites
80.3%
80.1%
76.2%
49.8%
41.5%
38.6%
24.3%
23.5%
7.0%


typographic queuing, embedded hotlinks, captions support graphics, captions

placed near to corresponding graphics, graphics support textual content,

highlighted key words, avoid jargon, slang, meanings of icons immediately

obvious, abbreviations, acronyms and initialisms, and graphics easy to read,

appeared on fewer than 500 web sites, while Information w/color also w/out

color, graphics used when messages are complex appeared on fewer than 200

visited sites. The message design element metaphors make unfamiliar familiar

was used on only 41 (7%) of 588 sites.

Research Question 2

Research Question 2-Which message design elements are CVB web

sites using effectively to deliver destination messages?-was also answered

using descriptive statistics. Effectiveness of design element use was

operationalized by rating each item on each CVB web site using a four-point

Likert-type scale, with 1 = poor, 2 = fair, 3 = good, and 4 = excellent. The mean

rating scores and standard deviations are reported in Table 4-6










Table 4-6. Ratings of message design element use on CVB web sites.

Number Mean Standard
Message Design Element o Ses cre Deviation
of Sites ScoreDeviation
familiar font 582 3.89 0.40
running text lower case 573 3.88 0.47
avoid jargon, slang 448 3.80 0.60
abbreviations, acronyms and initialisms 244 3.77 0.63
legible font 581 3.70 0.66
consistent color scheme 581 3.64 0.73
realistic colors used in graphics 570 3.62 0.82
color media preferable to black & white 572 3.61 0.90
text distinct segments 567 3.47 0.92
leading/interlinearspacing 575 3.46 0.94
short blocks text 567 3.42 0.95
picture overload 567 3.39 1.04
metaphors make unfamiliar familiar 41 3.34 1.06
picture is technically acceptable 564 3.32 0.91
multiple type styles 583 3.30 0.84
well organized text 574 3.29 0.97
picture fits in area 562 3.29 1.01
column width 574 3.21 1.02
font size appropriate 583 3.20 0.82
explicit text 572 3.20 0.93
picture fits with other pictures 513 3.20 0.98
picture depiction accurate 559 3.15 1.02
white space use 585 3.08 1.06
lines text 8 to 10 words 569 3.06 1.06
headings identify subject matter 578 3.03 0.98
graphic overview scrolling 565 3.02 1.12
picture is aesthetically satisfactory 564 2.85 1.07
loss text overview 584 2.79 1.10
meanings of icons immediately obvious 293 2.58 1.32
favorable first impression 579 2.55 1.02
graphics gain attention 565 2.51 1.08
graphics support textual content 472 2.50 1.16
headings attract attention 580 2.37 0.82
Information w/color also w/out color 143 2.27 1.21
graphics easy to read 227 2.25 1.21
typographic queuing 497 2.23 1.09









Table 4-6. Continued.

Number Mean Standard
Message Design Element N b Mean Standard
Message Design Element of Sites Score Deviation
macrosignals-underline, italics, bold 498 2.21 1.09
same info different formats 518 2.20 1.18
imbedded hotlinks for key terms 495 2.09 1.31
content generalized to specific 555 1.99 1.29
captions support graphics 487 1.70 1.16
captions placed near to corresponding graphics 485 1.70 1.19
graphics used when messages are complex 138 1.58 1.09
highlighted key words 471 1.37 0.87


Research Question 3

Research question 3-How many underlying instructional message design

conditions that enable desired learning outcomes are present on CVB web

sites?-was answered through exploratory factor analysis of the message design

variables to illuminate the conditions, or underlying factors, that explain the

shared variance of a total number of Likert-type variables. A principal component

analysis using varimax rotation with Kaiser normalization was conducted and the

results are presented in Table 4-7. Nine factors, explaining 64.7% of the

variance, were illuminated by the factor analysis. The item headings attract

attention did not load on any factor and was removed. One of the factors was a

single-item factor (content general to specific) and was removed from further

factor analysis, leaving eight factors that explained 61.4% of the variance.

The default setting for factor retention in most statistical software is to

retain all factors with eigenvalues over 1.0, although that may be the least

desirable method for determining the number of factors to retain (Costello &

Osborne, 2005). A scree test, widely utilized as a better method to determine







83


factor retention, is an examination of the plot of eigenvalues to determine the

natural break point where the curve flattens. The number of data points above

the break is recommended as the number of factors to retain (Costello &

Osborne). A scree plot of the factor loadings for this data set suggested

retention of four dimensions in the final model (Figure 4-1).



7-

6-

5-




.93-
LU

2-

1 -


0-
1 2 3 4 5 6 7 8 9 10 11 1213 14 15 16 1718 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Component Number

Figure 4-1. Scree plot of factor loadings.



Based on the scree test, four factors were selected for retention. Factor 1

was graphic presentation, which consisted of eight items and explained 21.7% of

the variance. Factor 2 was text presentation, consisting of 5 items and explaining

11.7% of the variance. Factors 3 and 4 each consisted of two items. In

exploratory factor analysis, the decision to retain factors with as few as two items

may be appropriate (StatSoft, 2006). In this study, both factor 3 and factor 4









contained highly correlated items with strong factor loadings, which subsequently

proved to be highly reliable measures, as well (Table 4-7). That evidence, taken

together with the scree test recommendation of four factors retained, justified

keeping factor 3 (visual text) and factor 4 (graphic/text support) in the final model.

The factor visual text explained 6.48% of the variance, while the factor

graphic/text support explained 4.04% of the variance.

The reliability of the scale was ascertained to indicate the precision of the

measurement or, more specifically, to assay how large the errors of

measurement were likely to be. The reliability coefficient (alpha) ranges between

0 and 1, where 1 denotes 100% error-free measurement. Table 4-7 presents the

reliability of the sub-scales. The four factors selected for retention each exhibited

acceptably high reliability. Graphic presentation (alpha = .89), text presentation

(alpha = .75), visual text (alpha = .84) and graphic/text support (alpha = .97) each

proved to be reliable measures. The remaining four factors each had a Cronbach

alpha of less than .7 and cannot be considered reliable measures.

The item short blocks of text loaded on both factor 2 and factor 3. Short

blocks of text is conceptually more aligned with the factor 2 attributes of

organization and segmentation than the factor 3 attributes of line length/column

width. Additionally, retaining short blocks of text on factor 2 raised the alpha from

.70 to .75, while removing short blocks of text from factor 3 raised the factor 3

alpha from .75 to .84, indicating greater reliability in each case. Retention or

exclusion of items in exploratory factor analysis requires judgment to produce the

cleanest fit of factors (Costello & Osborne, 2005). In this case, the conceptual











Table 4-7. Factor analysis and reliability of the message design variables.

Cron- Alpha if Std
Factor Eigen- % of Cron- Alpha if Std
Message Design Domain Loading value Variance bach's ite Mean ev
alpha deleted
Factor 1 Graphic Presentation 6.94 21.70 .89 2.93 1.00
picture is aesthetically acceptable .846 .87 2.86 1.05
favorable first impression .820 .87 2.55 1.01
pictures gain attention .789 .87 2.52 1.06
picture fits with other pictures on
the page .737 .88 3.21 .92
picture is technically acceptable .716 .88 3.32 .90
picture depiction is accurate .686 .88 3.16 1.00
picture fits into given area .645 .89 3.30 .99
graphics support text content .593 .90 2.52 1.05
Factor 2 Text Presentation 3.75 11.73 .75 3.28 .92
headings ID the subject matter .681 .76 3.03 .97
explicit text .660 .73 3.20 .92
well organized text .646 .67 3.29 .96
text in visually distinct segments .629 .67 3.47 .90
short blocks of text .527 .70 3.42 .87
Factor 3 Visual Text 2.08 6.48 .84 3.14 1.04
lines of text in 8-10 words .873 3.06 1.06
column width appropriate .837 -- 3.21 1.01
Factor 4 Graphic/Text Support 1.29 4.04 .97 1.71 1.07
captions placed near graphics .938 -- 1.71 1.08
captions support graphics .933 -- 1.70 1.06
Factor 5 1.66 5.18 .64 3.23 1.29
loss of text overview .842 .50 2.79 1.10
picture loss of overview .752 .51 3.02 1.10
white space appropriate use .435 .63 3.08 1.56
picture overload .430 .63 3.38 1.02
Factor 6 1.50 4.70 .55 2.92 .95
hotlinks imbedded in text .685 .44 2.10 1.20
same information in different
formats .601 .48 2.20 1.11
typographic queuing .568 .47 2.23 1.00
color pictures preferable to black
& white .467 .50 3.61 .89
Factor 7 1.27 3.96 .59 3.64 .77
realistic colors used in graphics .632 -- 3.63 .81
consistent color scheme .590 -- 3.64 .72
Factor 8 1.06 3.31 .54 3.33 .88
multiple type styles .620 .55 3.30 .84
font size appropriate .568 .41 3.2 .82
leading / interlinear spacing .470 .33 3.46 .93
Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.
Rotation converged in 14 iterations.
*Factor 1 = graphic presentation; Factor 2 = text presentation; Factor 3 = visual text; Factor 4 = graphics/text
support. Cronbach alpha's are in bold. Factors 5 through 8 exhibited low reliability were not retained.









contribution and elevated alphas resulting from inclusion of the item short blocks

of text in factor 2, and its removal from factor 3, were sufficient to justify those

actions. The alpha for the subscales graphic presentation, text presentation,

visual text, and graphic/text support was greater than .7, indicating that they were

a reasonably good measure of the multi-dimensional construct.

Research Question 4

To answer Research Question 4-Do regional or geographic differences

exist in the use of message design elements on CVB web sites?-an analysis of

variance by region was conducted. Significant differences were found between

world regions on the use of nine message design variables: font size appropriate,

picture loss of overview, pictures gain attention, favorable first impression, picture

depiction is accurate, picture is aesthetically acceptable, picture fits into given

area, picture fits with other pictures on the page, and consistent color scheme. A

Bonferroni corrected post-hoc test revealed that CVB web sites from Canada

were significantly different from those in three other regions on six message

design items. Oceania/South Pacific web sites also exhibited differences with

those from three other regions on six items. In these comparisons, Canada's

means were lower than the other regions. Conversely, the means for

Oceania/South Pacific were higher than the other regions. Table 4-8 includes all

the significant regional differences.

The analysis of variance revealed a significant group difference on the

factor graphic presentation, F(1, 587) = 3.319, p = .001. No significant

differences were found for the factors text presentation, F(1, 587) = 1.236, p =










.275; visual text, F(1, 587) = .968, p = .460; or graphic/text support, F(1, 587) =

1.499, p = .154. As depicted in Table 4-9, Bonferroni post hoc test indicated

significant differences on the factor graphic presentation between Canada and

six world regions (Africa, Central/South America, Europe, Caribbean, United

States and Oceania/South Pacific). There were no significant differences

between world regions for the message design factors graphic/text support, text

presentation, or visual text.



Table 4-8. Message design elements exhibiting significant differences between
regions.

95%
Message World World
Mesae ordMean WlMean Sig Confidence
Design Element Region Region en
Interval
consistent color Oceania/South
Asia 3.30 3.94 0.02 -1.22 -0.05
scheme Pacific
Sf Africa 3.05 0.03 -1.79 -0.05
favorable first
Canada 2.13 Oceania/South
impression anada 213 Oceania/South 3.03 0.00 -1.62 -0.18
Pacific
pictures gain Canada 2.06Oceania/South 3.00 0.00 -1.70 -0.18
attention Pacific
font size
t size Canada 2.91 United States 3.33 0.02 -0.80 -0.03
appropriate
picture is Africa 3.35 0.04 -1.81 -0.01
aesthetically Canada 2.44 Oceania/South 3.3 -. .
3.23 0.03 -1.53 -0.04
acceptable Pacific
SAfrica 3.59 0.04 -1.74 -0.02
picture depiction
picture depiction Canada 2.72 Oceania/South
is accurate 3.48 0.02 -1.47 -0.05
Pacific
picture fits into Canada 2.93 United States 3.41 0.04 -0.95 -0.01
given area
picture loss of Europe 2.64 United States 3.25 0.00 -1.01 -0.22
overview
consistent color Middle 15 Oceania/South 4 004 -154 -0
scheme3.15 3.94 0.04 -1.54 -0.02Pacific
scheme East Pacific










Table 4-9. Graphic presentation significant difference by region.

World Region mean Sig.
Canada (mean = 2.54) Africa 3.29 .00*
Oceana/South Pacific 3.18 .02*
Central/South America 3.08 .04*
Caribbean 3.06 .05*
Europe 2.96 .03*
United States 2.94 .01*
Asia 2.89 .49
Middle East 2.69 .99
* The mean difference is significant at the .05 level.


Other Variables Related to the Success of Destination Message Delivery

The nature of several important web site features did not permit

measurement as scale items. Download time, readability, and legibility of web

sites were analyzed separately, using software expressly designed to examine

these web site elements.

Download time for each web site was measured and means were

calculated for regions. Although the means of the nine world regions were not

significantly different, African web sites delivered the fastest download time

(Table 4-10) followed by Central/South America and the Caribbean. The six

remaining regions all had download times in excess of one second.

Readability was measured using the Flesch-Kincaid Grade Level

readability scale. The Middle-East mean score was 5.79, indicating that between

five and six years of schooling would be required to understand the textual

content on Middle-Eastern web sites. Mean scores for Canada, Africa, and Asia

indicated that between six and seven years of education were necessary, while










Table 4-10. Broadband download times by world

Std.
Region N S
Deviation
Canada 55 0.65
Middle East 13 1.02
Oceania/South Pacific 30 0.89
Asia 29 1.17
Europe 103 0.91
United States 270 0.82
Caribbean 30 0.55
Central/South America 24 0.70
Africa 18 0.63
mean 0.83


region.

Std.
Error
0.09
0.28
0.16
0.22
0.09
0.05
0.10
0.14
0.15
0.03


Mean
(in seconds)
1.38
1.33
1.27
1.26
1.22
1.18
0.95
0.94
0.87
1.19


scores for the United States and all other regions signaled that over seven years

of education would be required (Table 4-11).

There were no significant differences between world regions for the

legibility of the text and headings. Each world region achieved an acceptable

mean (Table 4-12) score for legible text, where scores closer to 1 indicate

greater legibility than scores closer to 2. African web sites scored the highest for

legible text, text/background color difference, and text brightness difference.


Table 4-11. Flesch-Kincaid grade level readability scores by region.

Region N SD Mean
Central/South America 24 3.42 7.49
Caribbean 30 1.72 7.19
Oceania/South Pacific 31 1.63 7.17
Europe 105 2.16 7.09
United States 272 4.06 7.08
Asia 30 2.71 6.97
Africa 16 1.88 6.93
Canada 55 2.16 6.81
Middle East 13 2.80 5.79
mean 3.24 7.05










Table 4-12. Legibility of text by region.


text color text brightness
Region N legible text
difference* difference**
Africa 17 1.06 652.18 221.00
Canada 55 1.13 639.42 211.82
Caribbean 29 1.14 644.52 220.62
Middle East 13 1.15 638.46 218.08
United States 276 1.15 651.74 219.88
Oceania/South Pacific 31 1.16 639.65 214.97
Asia 30 1.17 630.20 212.97
Central/South America 25 1.20 613.56 206.68
Europe 108 1.22 619.22 207.37
mean 1.16 640.54 215.66
*scores >500 are acceptable
**scores >125 are acceptable

The textual content was assessed for three types of color vision

impairment. There were no significant differences found between world regions

for color vision impairment, with each of the nine world regions achieving

acceptably high scores (Table 4-13).


Table 4-13. Color vision impairment mean scores for text by region.

Rion N text text text
Region N
protanopia deuteranopia tritanopia
Asia 30 1.17 1.17 1.13
Africa 17 1.06 1.12 1.06
Central/South America 25 1.20 1.16 1.16
Canada 55 1.18 1.18 1.20
Europe 108 1.24 1.24 1.24
Caribbean 29 1.14 1.21 1.10
Middle East 13 1.15 1.15 1.15
United States 277 1.16 1.17 1.17
Oceania/South Pacific 31 1.19 1.19 1.16
mean 1.18 1.18 1.17


The overall legibility of headings appeared to be acceptable for each world

regions, although four regions had unacceptably low scores (below 500) for









heading color difference (Table 4-14). There were, however, significant

differences found between text and heading color differences overall (Table 4-

15), and text and heading brightness differences overall (Table 4-16).


Table 4-14. Legibility of headings by region.

heading
legible heading color heading
Region N brightness
Region N headings difference* br
difference**
Asia 30 1.50 497.13 168.97
Africa 17 1.47 484.00 166.47
Central/South America 25 1.36 533.16 176.88
Canada 54 1.56 493.91 165.39
Europe 107 1.37 549.47 186.61
Caribbean 30 1.40 511.97 174.63
Middle East 13 1.46 485.46 159.62
United States 272 1.42 522.12 178.65
Oceania/South Pacific 31 1.48 504.84 173.13
mean 1.43 520.27 176.99
*scores >500 are acceptable (italics indicate unacceptable scores)
**scores >125 are acceptable


Table 4-15. Significant difference between text and heading color difference.

Variable t df Sig. (2- Mean 95% Cl 95% Cl
tailed) Difference Lower Upper
text color difference 23.81 583 .000* 140.54 128.94 152.13
heading color difference 2.94 574 .003* 20.27 6.74 33.80
*p is significant at .05; Test value = 500


Table 4-16. Significant difference between text and heading brightness
difference.

Variable t df Sig. (2. Mean 95% Cl 95% Cl
tailed) Difference Lower Upper
text brightness difference 46.89 583 .000* 90.66 86.9 94.5
heading brightness difference 21.50 573 .000* 51.99 47.24 56.74
*p is significant at .05; Test value = 125









The headings were assessed for three types of color vision impairment.

There were no significant differences found between world regions for color

vision impairment, with each of the nine world regions achieving acceptably high

scores (Table 4-17).


Table 4-17. Color vision impairment


Region


Asia
Africa
Central/South America
Canada
Europe
Caribbean
Middle East
United States
Oceania/South Pacific


N
30
17
25
54
107
30
13
272
31


mean


mean scores for headings by region.

Headings Headings Headings
Protanopia Deuteranopia Tritanopia
1.53 1.53 1.47
1.59 1.59 1.47
1.44 1.40 1.40
1.52 1.59 1.67
1.40 1.42 1.41
1.47 1.50 1.50
1.62 1.62 1.54
1.45 1.48 1.46
1.55 1.55 1.48
1.46 1.49 1.47


Color vision impairment for both text and headings was compared for

regional and overall differences. There were no significant differences found

between world regions for protanopia, deuteranopia or tritanopia. There were,

however significant differences found for the CVB web sites overall with respect

to differences in text vs. headings for protanopia (Table 4-18), deuteranopia

(Table 4-19) and tritanopia (Table 4-20).

Table 4-18. Text / headings difference: Protanopia.

Sig (2- Mean 95% Cl 95% Cl
Variable T df tailed) Difference Lower Upper

Protanopia (text) 11.24 584 .00* .18 .15 .21
Protanopia (headings) 22.40 578 .00* .46 .42 .51
*p is significant at .05; Test value = 1











Table 4-19. Text / headings difference: Deuteranopia.

Sig (2- Mean 95% Cl 95% Cl
Variable T df tailed) Difference Lower Upper

Deuteranopia (text) 11.43 584 .00* .18 .15 .21
Deuteranopia (headings) 23.59 578 .00* .49 .45 .53
*p is significant at .05; Test value = 1


Table 4-20. Text / headings difference: Tritanopia.

Sig (2- Mean 95% Cl 95% Cl
Variable T df tailed) Difference Lower Upper

Tritanopia (text) 11.11 584 .00* .17 .14 .21
Tritanopia (headings) 22.63 578 .00* .47 .43 .51
*p is significant at .05; Test value = 1


Research Question 5

To answer Research Question 5-What web page elements do web

design professionals believe to be most important for destination message

delivery?-a Delphi study using an international panel of web page design

experts was conducted. After being assigned fictitious names, panelists were

asked to ascribe importance ratings to a list of instructional message design

elements, and then discuss their item rankings in an effort to reach consensus.

Items were scored on a four point scale where 1 =not very important,

2=somewhat important, 3= very important, and 4=critically important. The final

item importance rankings (Table 4-21) illuminated areas of agreement, and

disagreement among the panel members. Of the 41 web page items under

consideration, unanimous consensus was reached for eight items.









Table 4-21. Delphi panel importance ratings of message design elements.

Mr. Mr. Mr. Mr.
Text / Graphics Elements Bron Rd Be Pk Mean
Brown Red Blue Pink


Whitespace (appropriate use)
Text color/background contrast &
brightness
Favorable first impression
Legible font (san-serif)
Font size appropriate
Graphics (charts) easy to read
Picture is aesthetically satisfactory
Picture fits into the given area
Consistent color scheme
Procurement (download) time
Running text in lowercase letters
Visuals & graphics gain attention
Picture is technically acceptable
Meanings of icons immediately obvious
Information w/color also w/out color
Headings attract attention
Familiar font
Embedded hotlinks
Use of readable language
Well organized text
Explicit text
Graphics support and augment textual
content
Captions placed nearer, rather than
farther, to corresponding graphics
Headings identify the subject matter
Leading / Interlinear spacing
(appropriate use)
Multiple type styles (appropriate use)
Captions support graphics
Picture depiction is accurate
Graphics used when messages are
complex
Text in visually distinct segments
Short blocks of text
Minimum loss of overview (scrolling)


4 4 4 4 4.00


4.00
4.00
3.75
3.75
3.75
3.50
3.50
3.50
3.50
3.25
3.25
3.25
3.25
3.25
3.00
3.00
3.00
3.00
3.00
3.00


2 3 3 4 3.00


3.00
2.75

2.75
2.75
2.75
2.75

2.75
2.50
2.50
2.50









Table 4-21. Continued.

Text / Graphics Elements Brown Red Blue Pink Mean
Avoid jargon, slang and other
unfamiliar words 2 4 3 1 2.50
Color media preferable to black-and-
white 2 3 1 4 2.50
Minimum loss of overview (scrolling) 2 2 2 2 2.00
Same information is delivered in
different formats 3 2 2 1 2.00
Picture fits in with the other pictures on
the page 3 1 3 1 2.00
Column width 2 2 2 1 1.75
Abbreviations, acronyms and initialisms
(appropriate use) 2 1 3 1 1.75
Picture overload 2 2 3 1.75
Lines of text in 8 to 10 words 2 2 1 1 1.50
Provide metaphors or quotes to make
the unfamiliar familiar 1 1 3 1 1.50
Content generalized to specific 2 2 1 1 1.50
Realistic colors used in graphics 1 1 2 2 1.50
Highlighted key words 1 1 2 1 1.25
Macrosignals (underline, italics, bold) 1 1 1 1 1.00


All of the panel members expressed a belief that whitespace, text

color/background contrast and brightness, and favorable first impression were

"critically important" design elements. Panel members were also united in casting

loss of overview as somewhat important, and macrosignals as "not very

important". Headings attract attention, well-organized text, and explicit text were

deemed very, although not critically, important. Near-unanimous agreement of

high importance was reached on seven additional items, legible font, font size

appropriate, graphics/charts easy to read, picture is aesthetically satisfactory,

pictures fits into the given area, consistent color scheme, and procurement









(download time). Of lesser importance were the variables running text in

lowercase letters, visuals and graphics gain attention, picture is technically

acceptable, meanings of icons immediately obvious, Information w/color also

w/out color, headings attract attention, familiar font, embedded hotlinks, use of

readable language, well organized text, explicit text, graphics support and

augment textual content, and captions placed nearer, rather than farther, to

corresponding graphics. Considered to be somewhat important were the

variables headings identify the subject matter, leading interlinearr spacing,

multiple type styles, captions support graphics, picture, depiction is accurate,

graphics used when messages are complex, text in visually distinct segments,

short blocks of text, minimum loss of overview (scrolling), avoid jargon, slang and

other unfamiliar words, and color media preferable to black-and-white. Delphi

panel members considered column width, abbreviations, acronyms and

initialisms, picture overload, lines of text in 8 to 10 words, provide metaphors or

quotes to make the unfamiliar familiar, content generalized to specific, realistic

colors used in graphics, highlighted key words, and the aforementioned

macrosignals to be of relatively minimal importance with respect to the other web

page elements under discussion.

Summary

Data from 588 CVB web sites were analyzed to examine the extent of the

use of instructional message design principles. While many message design

items were nearly ubiquitous, others were little used. Three items appeared on

less than 25% of the CVB sites; six appeared on less than 50%. All other









message design items, however, appeared on over 75% of the web sites visited.

Factor analysis produced a four-dimension model of web design factors that

describe the construct of an instructional environment for destination message

design. The four dimensions were "graphic presentation," "text presentation,"

"graphic/text support," and "visual text." There was a significant difference found

between world regions for the factor "graphic presentation." Finally, experts in

web design were called on to determine which variables they considered

important in web design. In general, the Delphi panel considered the variables

that dealt with the visual aspects of text and graphics to be the most important.

Chapter 5 presents an interpretation of these results and implications for both

practice and further research.














CHAPTER 5
DISCUSSION

In Chapter 4, the results of the data analysis were presented. Chapter 5

provides a discussion of these analyses, linkages to theory and findings from

previous research, and implications for practice and future research.

Discussion of the Results

This study examined the use of instructional message design principles in

the design of CVB web sites around the world. Specifically, the nature and

frequency of use was evaluated, and web sites were compared to detect regional

differences. Web design experts were consulted to determine which elements

were considered important to design professionals. The following sections

provide an in-depth discussion of the results.

Appearance of Message Design Elements

For Research Question 1- Which elements of message design appear

on CVB web sites?-descriptive statistics revealed that, to a varying degree,

each of the 588 CVB web sites contained some of the message design elements

under review in this study (Table 4-5). The most pervasive message design

practices were picture fits with other pictures, same information presented in

different formats, content written from generalized to specific, picture depiction is

accurate, picture fits into given area, picture is aesthetically satisfactory, picture

is technically acceptable, graphic overview (scrolling), graphics gain attention,









text in visually distinct segments, short blocks of text, picture overload, lines of

text in 8 to 10 words, realistic colors used in graphics, explicit text, color media

preferable to B&W, broadband (download speed), running text lower case letters,

column width, well organized text, leading interlinearr spacing, headings identify

the subject matter, favorable first impression, legible headings, headings attract

attention, legible font, consistent color scheme, familiar font, font size

appropriate, multiple type styles, loss of text overview, legible text, and

whitespace (appropriate use), all appearing on more than 500 web sites.

Many message design (30) items were nearly ubiquitous, appearing on

more than 95% of the web sites assessed. Six of these, familiar font, font size

appropriate, multiple type styles, loss of text overview, legible text, and

appropriate use of whitespace appeared on 99% of the CVB sites visited. All

(100%) of CVB web sites use at least some of the instructional message design

practices recommended to ease the cognitive load of web site visitors. This result

is not surprising, as many message design elements had their genesis in print

media and are familiar to both print and web authors alike, albeit not as methods

to reduce cognitive load but as style guides for best writing practices.

Although writing for the web is different from writing for print media (e.g.,

scannable text, etc.) many of the same conventions developed for print have

been adapted for electronic media. In each case, content is created for readers

or viewers who have become accustomed to familiar presentation styles, whether

in print or on the web. Most graphic designers follow traditional typographic rules

for constructing text and displaying graphics. As a result, readers have become






100


accustomed to similar standards in typesetting, a familiarity that has become a

reader expectation (Keller & Burkman, 1993).

Participants in the Delphi study, with their emphasis on visual impact,

reiterated the importance of many of these commonly used elements. For

example, several participants emphasized the necessity of a "favorable first

impression," or as one participant expressed, "I don't stay if I get a bad first

impression. If they can't do that right, what's next? It just gets worse." Specific

common elements were considered essential, as well. For example, "whitespace

is something everyone notices without knowing it. It registers quietly. Your

average user never notices. But it affects them anyway." Many of the other

elements that were most commonly found on CVB sites were also rated as

"critically important" or "very important" by the Delphi panelists.

Some message design elements appeared on relatively few sites. The

use of metaphors to make the unfamiliar familiar (appearing on only 41 of 588

CVB sites), information presented with color also presented without color (143

sites), and use of graphics when messages were complex (138 sites) were the

least used message design techniques. Interestingly, these three practices all

share a redundancy component, as each of these three variables deals with

providing a secondary mechanism for content delivery. Winne (1995) suggested

that learning improves when the same information is delivered in different

formats, but the results of this study indicate that the utilization of multiple

formats for web content has been underused on CVB web sites.









Members of the Delphi panel considered two of these little-used elements

as unimportant: metaphors to make the unfamiliar familiar and information

presented with color also presented without color. The third element, use of

graphics when messages were complex, which appeared on only 138 of the 588

sites, was considered "very important" by three of the four panel members and

"somewhat important" by the fourth. There are several possible explanations for

the failure of many CVB site designers to include this important element. Sites

with no complex messages would have no reason to include graphics for this

purpose; there was no distinction on the scale between sites that did not employ

graphics when messages were complex and those that had no complex

messages. A potentially problematic explanation for the omission of graphics

when messages were complex is that designers may not have considered their

messages to be complex. This may occur because the web designer may be

completely disconnected from the content of the site; that is, the web designer

may be more concerned with making the site look good and making it functional

than making it an effective conveyor of complex messages. This assumption is

supported by the responses from the Delphi panel members.

A critical function of CVB web sites is to convey information in a way that

will promote the user's retention. According to cognitive load theorists (Sweller,

1994; Sweller, Van Merrienboer, & Paas, 1998), this can be most efficiently

accomplished by reducing the demands on working memory. Content should be

presented to focus the user's working memory on relevant information and to

avoid taxing it with irrelevant or distracting features. The widespread use of






102


numerous effective message design elements indicates that CVB web designers

are at least somewhat aware of the need to support users' retention of

information.

The responses of the Delphi panelists indicate that web professionals

have accepted many of these elements as essential. However, even several of

the elements that were considered unimportant by the Delphi panelists were

found on most of the web sites examined. For example, macrosignals

(underline, italics, bold) were considered the least important by the panelists, yet

84.7% of the sites examined employed this element. Similarly, highlighted key

words were used by 80.1 % of the CVB web sites.

Effectiveness of Message Design Elements on CVB Web Sites

Research Question 2-Which message design elements do CVB web

sites employ effectively to deliver destination messages?-was answered using

the mean scores from the raters' checklists. Although, upon examination of the

frequency data, it became evident that the message design elements were

commonly used, the quality of their use was quite variable. The most striking

aspect of the variability in quality was that the mean scores for visual impact-

related items tended to be high, but the mean scores for content delivery-related

items tended to be low.

CVB web sites' use of text (e.g., font size, familiarity, and legibility, short

blocks of text, well organized text, distinct segments of text) and graphics (e.g.,

realistic colors, technical acceptability of pictures, use of white space) was

consistently rated very high (3.0 or higher). This indicates that CVB web sites






103


are succeeding in the area of visual impact. In contrast, items related to making

the destination message accessible and understandable tended to be rated

lower. For example, items such as captions support graphics, macrosignals

(bold, italics, underline), imbedded hotlinks for key items, and content

generalized to specific were all rated below 2.25. While CVB sites tend to be

visually appealing, the poor ratings on these items indicate that they may be

failing in their primary mission of conveying a destination message.

This finding is consistent with the Delphi panelists' emphasis of the

importance of visual impact. The panelists rated most of the content delivery-

related items as unimportant or only somewhat important. This disconnect

between the focus of web designers and the mission of the CVB is problematic.

Factor Analysis

Research Question 3-How many underlying instructional message

design conditions that enable desired learning outcomes are present on CVB

web sites?-was answered through the data reduction technique of factor

analysis of the scale variables to illuminate the conditions or factors that explain

the pattern of correlations within the CVB dataset. Factor scores were saved as

new variables and used to interpret what the factors represented conceptually.

Four factors were retained in the final model. The first factor, "graphic

presentation" contained the items picture is aesthetically acceptable, favorable

first impression, graphics gain attention, picture fits with other pictures on the

page, picture is technically acceptable, picture depiction is accurate, picture fits

into given area, and graphics support the text. All of the items comprising the






104


factor "graphic presentation" pertain to the visual display of graphic content. The

alpha for this factor was quite high (alpha = .89), indicating that the above items

were a reliable measure of the presentation of graphic content. This factor

represents an important aspect of message design, because recall of pictures is

better than recall of text (Gagne & Rower, 1969) and visuals and graphic images

are also perceived more rapidly than texts (Fleming & Levie, 1978). This factor is

also critical in presenting a favorable first impression. Learners are more likely to

feel confident and motivated about materials that convey a favorable first

impression than those that do not. Members of the Delphi panel agreed on the

importance of creating a favorable first impression, calling it a "critical" concern

for designers.

An often-repeated opinion on the function of visual images concerns

getting attention and maintaining attention (Pettersson, 1998). Like the items that

comprise "graphic presentation", most motivational graphic principles relate to

gaining and maintaining attention (Keller & Burkman, 1993). Witt (1981)

explained that, although pictures and graphics may not carry the main weight of a

message, they are useful in attracting the direct attention of the learner.

Attention is the mechanism that determines which stimuli will be engaged in and

which stimuli will be ignored (Carlson, 1993). Mayer (1993) stated that for

learning to occur, the learner must first pay attention to the relevant information.

To accomplish this goal, instructional designers must effectively draw the

learner's attention to the relevant information.






105


The Delphi panelists were somewhat concerned with graphics, as they

rated several elements related to graphics (e.g., graphics (charts) easy to read,

picture is aesthetically satisfactory, picture fits into the given area) as "very

important" or "critically important." Interestingly, the Delphi panelists were less

concerned with whether the graphics contributed substantively to the message of

the site. They rated several items related to the content of the picture (e.g.,

graphics used when messages are complex, captions support graphics, picture

fits in with the other pictures on the page) as less important. The focus of these

web design experts was very clearly on the visual impact and functionality of

graphics, not on the message delivery aspects. For example, one panelist

explained, "People don't read most captions. Unless I need a caption to identify

someone, I avoid using them. As a rule." Again, the disconnect between web

designers' focus on visual impact and functionality of a site and the CVB's

mission of conveying a destination message may be at odds.

The appropriate use of textual features can increase attention and

influence the motivation to learn, (Keller & Burkman, 1993). The second factor,

"text presentation," was composed of five message design items that pertained to

the visual display of text. These factors were headings identify the subject matter,

explicit text, well-organized text, text in visually distinct segments, and short

blocks of text. This factor was important because the formatting of text-based

content can determine whether it is perceived as easy or difficult to use (Keller &

Burkman). Relatively short blocks of text and concise messages tend to suggest

fairly easy reading, and that initial easy-to-read perception builds confidence and






106


maintains attention (Keller & Burkman). Designers can also create the

perception of easy reading by dividing lengthy blocks of text into visually distinct

segments (Keller & Burkman; Smith & Orr, 1985; White 1987). The third factor,

"visual text" (alpha = .77), pertains to the layout of text on the page and was

composed of three items lines of text in 8-10 words, column width appropriate,

and short blocks of text. Research indicates that lines of text should be

composed of eight to ten words in 12-point font to make text easier to read,

maintain learner attention and increase confidence, and that lines of text should

also be presented in a fairly open, rather than constrained display (Keller &

Burkman, 1993).

The Delphi panelists rated many of the items related to "text presentation"

and "visual text" as "critically important" or "very important." In particular, items

related to the font size and legibility and the contrast between text and

background colors were considered critical. Again, these items relate to the

visual impact of the web site, not to its content. In contrast, items such as

organizing content from generalized to specific and using macrosignals (bold,

italics, underline) or highlighting of key words to emphasize important concepts

were considered not important at all. The panelists considered text elements

such as text in visually distinct segments and short blocks of text only "somewhat

important." These elements contribute significantly to the readability of text and,

therefore, to destination message delivery.

The final factor, "graphic/text support," was composed of two items, each

pertaining to the use of captions (captions support graphics, captions placed near






107


to corresponding graphics) within web page design. The high alpha (.97) for this

factor is not surprising, as a high correlation between the two "caption" items

should be expected. This finding is helpful for web message designers, as

physical integration of visual and verbal content can decrease cognitive load and

increase learning (Chandler & Sweller, 1991). Sound design ensures that

learning is enhanced when text is placed nearer, rather than farther, to

corresponding pictures and graphics (Mayer, 1993). When both text and graphics

convey exactly the same information, users may become bored or distracted.

Text and graphics should be complementary, offering slightly different views of

related information to promote learning (McFarland, 1995). As previously

mentioned, the Delphi panelists were not particularly concerned with the use of

captions. In fact, one member of the panel avoids using them altogether.

Regional Differences in the Use of Message Design Elements

Research Question 4-Do regional or geographic differences exist in the

use of message design elements on CVB web sites?-was answered by an

analysis of variance of the nine world regions contained in the CVB data set.

Given the uneven diffusion of the Internet, regional differences were expected.

Tufte (1990) noted that the principles of information design are universal

and are not aligned with any particular language or culture. Still, there was a

significant difference found between world regions on the use of nine message

design variables: font size appropriate, picture loss of overview, pictures gain

attention, favorable first impression, picture depiction is accurate, picture is

aesthetically acceptable, picture fits into given area, picture fits with other






108


pictures on the page, and consistent color scheme. The differences were related

to lower scores of sites in the Canada region and higher scores of sites in the

Oceania/South Pacific Region. These differences may be due to the differences

in the degree of reliance on tourism in the respective regions. That is, the

Oceania/South Pacific region relies more on tourism, which dominates the

economies of many countries in the region. The remote locations of many of the

island nations create more reliance on the Internet to communicate with potential

visitors than some of the other regions. In comparison to islands in

Oceania/South Pacific, tourism is a less integral part of the Canadian economy.

Also, with the US as a primary source of tourists, Canadian CVBs have other

outlets available for communicating their destination messages.

A further analysis of variance on the four factor subscales applied to the

nine world regions revealed a significant difference on one of the factors,

"graphic presentation." This difference was between Canada and other world

regions. Given the lesser importance of tourism to the economy in Canada

compared with some other world regions, it would logically follow the investment

in CVBs and, in particular, CVB web sites in Canada would be less. Graphic

presentation is the factor most dependent on professional designers. The other

factors, which are more closely aligned with well-established print conventions,

consist of elements that can be generated by designers with less expertise. High

quality graphic design requires more skill. Therefore, one explanation for the

differences on this factor could be differences in CVB budget allocations to web

design in Canada. Additional study would be required to confirm or refute this






109


explanation. There were no significant differences found between world regions

for the remaining three subscales.

Other Variables Related to the Success of Destination Message Delivery

The web site visitors' exposure to site content can be influenced by a

plethora of technical issues, such as the speed of the ISP, computer settings and

capabilities, and web browser preferences and settings. The combination of

these features defines the technical environment that affects the display of

information available to site visitors (Hodkinson & Kiel, 2003). Several important

web site features were not measured as scale items. Download time, readability,

and legibility of web sites were measured using specialized software useful for

examining such elements. Scores for these items were compared to detect any

regional differences.

Web user impatience with overlong waiting times is a problem that can, in

part, be solved by good design. Download time is a critical area of concern for

web users, and extended waiting times are not suffered gladly in the wired

environment. Research has indicated that tolerable waiting time for web users is

about two seconds, after which users begin to search for alternate sites (Fui-

Hoon Nah, 2004). As broadband download time for the CVB web sites in this

study averaged 1.19 seconds, there does not appear to be a widespread

problem in this area. However, 13.9% of the web sites measured reported

download times in excess of two seconds, and a few sites ranged a high as 4.8

seconds. It should be noted that download time is not wholly a function of web

site design and connection speed, and that download times for CVB web sites






110


were recorded as a "snapshot" of conditions on a particular day. For download

time, there were no significant differences found between the regions. This was

not unexpected, as the structure of the World Wide Web spans the regions, that

is, each region is not an autonomous web with self-contained systems

comparable to other world regions. Significant differences between web site

download time could be attributable to many factors, such as server speed or

total net traffic. As such, world region may not be a valid parameter with which to

assess download speeds.

Using readable language is an important message design consideration,

as readable language helps gain and maintain learner attention (Keller &

Burkman, 1993). Although readability typically refers to the ease with which the

language in text is read and understood (Chall, Bissex, Conrad, & Harris-

Sharpies, 1996), other aspects of text also contribute to its readability.

Readability was measured using the Flesch-Kincaid Grade Level

readability scale. Hull (2004) reports that clear and simple language is

recommended when designing for readers with different reading levels. Keller

and Burkman (1993) recommended that designers consider their target learner's

reading skills and interest in reading when designing text.

There were no significant differences between world regions in the

readability of the web sites visited. The Middle-East mean score was 5.79,

indicating that nearly six years of schooling would be required to understand the

textual content on Middle-Eastern web sites. Mean scores for Canada, Africa,

and Asia were slightly higher (between six and seven years of education), while









scores for the United States, Europe, Oceana/South Pacific, the Caribbean, and

central/South America were all somewhat over seven years. Hull (2004) noted

that the average resident of the United States had an eighth-grade reading level,

therefore, a Flesch-Kincaid score somewhat over seven would be appropriate for

the American target audience.

It should be noted that the Flesch-Kincaid Grade Level readability scale

scores all rendered content, including navigational items, which could skew the

results for a particular web site containing little text but a great deal of navigation.

The legibility of the text and headings on CVB web sites was measured as

a function of the foreground/background color contrast. There were no significant

differences between world regions for legibility of the text or headings. Although

the overall legibility of headings appeared to be acceptable for each of the world

regions, four regions had unacceptably low scores (below 500) for "heading color

difference" (Table 4-14). There were, however significant differences found

between text and heading color differences overall (Table 4-15), and text and

heading brightness differences overall (Table 4-16).

Additionally, the data revealed that just over 16% of the CVB sites visited

did not exhibit sufficiently legible text, and over 42% did not exhibit sufficiently

legible headings. CVB sites with such poor legibility scores can hardly be said to

be focused on Keller and Burkman's (1993) recommendations for maintaining

learner attention or increasing confidence. In particular, design of text and

headings that fail to meet legibility standards seems counterproductive in a

message design context, and may reflect an approach to design that seeks to






112


engage the eye at the expense of engaging the brain. During the Delphi study,

one of the panel members indicated that they would opt for less distinct text if it

blended better with the overall page design. Viewed from that perspective, it is

not surprising that text, and headings in particular, could be produced with less

emphasis on legibility than was optimal.

The text and heading foreground/background color contrast was also

assessed for three types of color vision impairment. There were no significant

differences found between world regions for text color vision impairment, with

each of the nine world regions achieving acceptably high scores (Table 4-14).

Overall, however, the presentation of text and headings for the visually impaired

varied greatly. For text, 17.7% of the web sites visited failed to provide adequate

color conditions for site visitors with Protanopia, while 18.2% failed for

Deuteranopia, and 17.3% failed for Tritanopia. For headings, the results were

much worse. For the CVB web sites visited, 45.7% failed to provide adequate

color conditions for visitors with Protanopia, 48.3 failed for Deuteranopia, and

46.3 failed for Tritanopia. Hartley (1987) observed that Deuteranopia (red-green)

color blindness is the most common, and is more widespread among men than

among women. Unfortunately, red and green are also used very often as key

colors in symbols and graphics (Pettersson, 1998). The high level of failure to

provide adequate color conditions on CVB web sites is a message design

problem that renders the site content inaccessible for many users.






113


Importance of Message Design Elements to Web Design Professionals

To answer Research Question 5-What web page elements do web

design professionals believe to be most important for destination message

delivery?-an online Delphi study was conducted using an international panel of

web site design experts. Given the ever-changing nature of the Internet and, in

particular, of web design practices, consultation with practicing experts in the

field was necessary to allow for more meaningful interpretation of the data and a

better understanding of the implications of this study for professional practice. In

addition to the relevant Delphi panel comments previously cited, a summary of

the entire Delphi study is provided in the following sections.

All of the Delphi panel members agreed that whitespace, text

color/background contrast & brightness, and favorable first impression were

"critically important" design elements. Bevan (1998) cautioned that unnecessary

whitespace may make scanning of the text more difficult, resulting in an

imposition on gaining attention. Delphi panel members, however, viewed

whitespace as "part of creating a favorable first impression" and was important as

"the background tying everything together." Panelists, however, thought that

creating a favorable first impression was more important, "by far." Although

united in their feeling that text color/background contrast & brightness was

critically important, panelists were less certain of its absolute necessity, stating

that "it's a tonal decision," and that "people will go with whatever looks better"

when opting for less distinct text if that is what fits better with overall design.






114


Items that rated very highly, although not unanimously critical, were legible

font, font size appropriate, graphics easy to read, picture is aesthetically

satisfactory, picture fits into given area, and consistent color scheme. Panelists

believed that a web site should have a "professional appearance", echoing

Kormier's (1999) statement that a beautifully designed, current, and informative

web site can establish a company's presence and enhance the identity of the

company.

Panelists thought that having a consistent color scheme was a key factor.

The panel reiterated that creating a favorable first impression was the most

critically important item, and that all of the visual elements were tied into that

objective. Keller and Burkman (1993) agreed, underlining the importance of

designing text that creates a favorable first impression, as a favorable first

impression enables the reader to scan the content and feel confident that they

can easily read and understand the message. A favorable first impression may

help to gain and keep attention and build confidence that readers can

successfully navigate and use the content. Gaining and keeping the attention of

the reader is also a useful method of reducing the imposition of nonlearning

events on working memory.

The next group of web page elements (in order of importance) includes:

running text in lowercase letters, headings attract attention, familiar font, visuals

& graphics gain attention, picture is technically acceptable, procurement

(download) time, meanings of icons immediately obvious, imbedded hotlinks, use

of readable language, well organized text, explicit text, minimum loss of overview






115


(scrolling), graphics support and augment textual content, and captions placed

nearer rather than farther to corresponding graphics. One panel member made

the distinction that scrolling up/down was acceptable, while scrolling from side-to-

side was not, claiming that "scrolling side-to-side is awful & should never

happen". Another panel member cautioned that "impact, navigation and content

are the critical design items", and that "we're splitting hairs on a lot of these",

reflecting the statements of several researchers: (a) Jung and Butler's (2000)

important characteristics of a successful web sites that included appearance,

ease of navigation, design, and content; (b) Bevan's (1998) statement that

successful web site development was contingent upon the developer's

awareness of graphic design and web usability; and (c) that most web site

developers do not consider aspects that are specific to education and the

teaching-learning process (Montilva, Sandia, & Barrios 2002).

Other graphic related items such as visuals and graphics gain attention,

graphics support and augment textual content, and captions placed nearer,

rather than farther, to corresponding graphics were all deemed to be important

for gaining attention. Interestingly, one panel member added that much of his

opinion/ratings were "subjective", and "I don't guarantee that my answers would

be the same tomorrow." He cautioned that design decisions had to be made "in

context", and warned that "all my scores are as a rule, but they all could change

in context".

Panel members were also united in casting loss of overview as somewhat

important, and macrosignals as "not very important". Lastly, a panel member






116


noted that "If we screwed up any of these items, very few would be catastrophic.

Whitespace was unanimously important, but if I didn't quite get that right, few

would notice".

In general, Delphi panel members considered design for the web to be an

almost wholly visual experience ("It's on a monitor!"). This viewpoint reflects the

findings of Rayman-Bacchus and Molina (2001), who noted that the success of

leading tourism sites suggested a competitive advantage for sites that were

visually appealing and information rich, and Pettersson (1999), who considered

that message design principles could be best utilized as guidelines for the display

of a visual messages. The experts believed that functionality and aesthetics were

the most essential web site features, and admitted that they seldom thought

about content delivery in the terms presented in this study. This finding is

particularly important, because the mission of CVBs is to deliver destination

messages (Gallarza et al., 2002). If content on a CVB web site is not acquired or

retained by the user because the web site designer is more concerned with

aesthetic elements than content, the web site will not perform its function in

helping the CVB achieve its mission.

Limitations and Delimitations of the Study

Although the investigator rigorously adhered to predetermined evaluation

criteria, inherent subjectivity by multiple raters in application of the measurement

instrument was unavoidable. Although interrater reliability was established within

acceptable limits for this study, a small degree of error variance between raters

was present in the final results.






117


The stratification of the sample into world regions was based on the

source site (tourismbureaus.com). The evolving nature of the Internet

subsequently affected the availability of some sites in the sample, producing

some "experimental mortality" or differential loss of cases due to web sites

dropping out of the study on a potentially non-random basis.

Because this study examined only CVB web sites, results may not be

generalized to commercial tourism sites. Generalization of the findings to other

types of informational sites may be similarly inappropriate.

The study of instructional message design was developed, primarily, in

western countries. Application of the tenets of message design to web sites

developed for non-western countries may present an imposition of cultural bias.

However, given that the study was conducted on only English-language CVB

sites, the presumption that the target audience is predominantly western may be

valid.

In a related vein, regional differences cannot be considered cultural

differences, because several of the regions included in the study are inhabited by

people from many cultures. Web design is conducted without respect to national

borders; that is, a web site representing a CVB may be created by a web

designer in a different part of the world. Finally, although the data have been

classified by region, regional classification does not imply regional causality.

Implications for Practice

Ultimately, the design of web sites for Convention and Visitors Bureaus

(CVBs) must consider the role of the CVB and the goals of the users. CVBs help






118


to generate tourism by providing information in the form of a clear and appealing

destination message.

This study revealed that CVB web sites vary widely in their use of effective

message design principles. Although numerous design elements appear to be

ubiquitous (e.g., headings, use of a consistent color schemes), others were rarely

found (e.g., use of metaphors).

This study yielded several important findings for CVB marketing personnel

and web site designers. A wide range of expertise exists among CVB web

designers, and an apparent bias in favor of visual- over content-related elements

exists among experts. One the tenets of any type of design is that form follows

function. If the function of a CVB web site is to deliver a destination message,

that function cannot be achieved when message design precepts related to

content are ignored.

An understanding of the essential features of instructional message

design would increase the likelihood that designers would employ these features.

Many sites exhibited features such as prominent but irrelevant animations or

other visual distractions, demonstrating a lack of understanding of the basic

elements of effective message design. Getting attention and maintaining

attention are two separate tasks. Maintaining attention is made more difficult by

the use of distracting elements such as advertising, pop-up windows, or music on

web sites, all of which increase cognitive load. As cognitive load theorists

suggest, maintaining attention is essential to promote retention of information.






119


Several recommendations for improving the design of CVB web sites

emerged. Web sites should support the user by reducing cognitive load. This

can be accomplished by making sites that are visually appealing, logically

organized, and considerate of the user by incorporating the principles of

instructional message design. Web designers should consider the elements of

effective message design as more important than the technical aspects of the

site. Although functionality is essential, the slick features many designers use in

an apparent attempt to look state-of-the-art can actually serve as distractions to

the user. Designers must remain focused on the purpose of the CVB. Finally,

CVBs should ensure that their destination message is conveyed in a way that

web site users will receive and retain the intended message.

As both the Internet communications and tourism industries grow, it is

likely that the design of CVB web sites will become an increasingly specialized

field. Specialists in this area should recognize the fundamental difference

between the purpose of CVB web sites and other sites. The purpose of a CVB

site is to increase tourism to an area by educating potential visitors about the

positive aspects of the destination. To accomplish this goal, implementation of

effective instructional message design principles is essential.

Implications for Future Research

Although this study answered important questions about the use of

message design principles on CVB web sites, it produced many questions, as

well. Some of these questions resulted from the factor analysis and others were

raised during the Delphi study.






120


A logical follow-up to the factor analysis would be to conduct a cluster

analysis to determine how individual sites clustered on the four factors. A cluster

analysis could enhance understanding of the issues contributing to the use of

effective message design. Regional or population variables may be important

within the clusters.

Is the effective use of message design principles related to characteristics

of the CVB? A study to examine this issue would consider factors such as the

size of the population represented by the CVB; that is, are web sites representing

countries more effective than those representing states, provinces, or

municipalities? In addition, is the effectiveness of the web site related to the

CVB's budget? Is the popularity of the destination related to the effectiveness of

the web site? Presenting an effective destination message is more critical for

destinations that rely more heavily on tourism for their economic vitality.

To control for the wide variability in the size (i.e., number of pages and

features) of web sites, this study was limited to examination of the lead page of

the CVB sites in the sample. A follow-up study to examine the functionality of

sites beyond the lead page is warranted. That is, further research is necessary

to determine the efficiency and ease of use of the navigation features, and the

ease with which the user can find desired information. The experts in the Delphi

study considered the functionality of sites to be of critical importance. Because

navigation was outside the scope of this study, a follow-up study on usability of

CVB web sites may be warranted. Similarly, additional research on the

accessibility of the sites would be useful.









The experts in the Delphi study were also concerned with aesthetics of

web sites. The items on the scale dealing with aesthetics were limited to

variables that made the site more readable and made features of the site more

easily distinguishable. The scale did not include items focused on how pleasing

a site was to the eye or how appealing the text or graphics were to the viewer.

For this reason, a follow-up study examining the subjective preferences of the

user may prove to be illuminating.

Future research should examine the use of features that may distract the

user. Many sites use animation or other eye-catching elements to attract

attention to particular features on the site. Many also include commercial

advertising, which is likely used to offset the costs of creating and hosting the

web site. A study of the relative effects of attractions versus distractions on CVB

sites could provide further insight into what makes a site effective.

Perhaps the most important direction for future research is to include

travelers in the rating of CVB web sites. A study that examined users' capacity to

retain information from sites exhibiting the various features included in this study

would be particularly valuable.














APPENDIX A
INSTRUCTIONAL MESSAGE DESIGN ELEMENTS
ASSESSED ON CVB WEB SITES

Text Elements


Reference


Loss of overview (scrolling)

Whitespace (appropriate
use)
Text in visually distinct
segments

Short blocks of text





Running text in lowercase
letters
Lines of text in 8 to 10 words


Macrosignals (underline,
italics, bold)
Text color/background
contrast & brightness


Typographic queuing


Headings attract attention


Weitl, Si3, Kammerl, & Freitag, 2002,
Dyson (2004).
Bevan, 1998, MacGregor & Lou (2004),
Shneiderman & Plaisant, 2005.
Keller & Burkman, 1993; Smith & Orr,
1985; White 1987; Allen & Eckols (1997),
MacGregor & Lou (2004).
Keller & Burkman, 1993; Allen & Eckols
(1997); McGovern, Norton & O'Dowd,
2002, Weitl, Si3, Kammerl & Freitag,
2002., Lee & Boling, 1999, MacGregor &
Lou (2004).
Henny, 1981; Poulton & Brown, 1968, Lee
& Boling, 1999.
Keller & Burkman, 1993, Lee & Boling,
1999, Dyson (2004), Shneiderman &
Plaisant, 2005.
Cisotto and Boscolo (1995), Lee & Boling,
1999.
Pettersson (1991), Lee & Boling, 1999,
Woodland & Szul (1999), MacGregor & Lou
(2004), Hall & Hanna (2004), Shneiderman
& Plaisant, 2005.
Pettersson (1991); Allen & Eckols (1997);
McGovern, Norton & O'Dowd, 2002, Lee &
Boling, 1999, MacGregor & Lou (2004).
Bevan, 1998, Cisotto and Boscolo (1995);
Allen & Eckols (1997); McGovern, Norton &
O'Dowd, 2002. Lee & Boling, 1999,
MacGregor & Lou (2004).


122


Item









Headings identify the subject
matter
Highlighted key words

Leading / Interlinear spacing
(appropriate use)
Column width

Abbreviations, acronyms and
initialisms (appropriate use)
Avoid jargon, slang and other
unfamiliar words
Provide metaphors or quotes
to make the unfamiliar
familiar
Same information is
delivered in different formats
Use of readable language


Well organized text

Explicit text

Content generalized to
specific
Familiar font


Legible font (san-serif)


Font size appropriate

Multiple type styles
(appropriate use)
Captions support graphics


McGovern, Norton & O'Dowd, 2002,
MacGregor & Lou (2004).
Bevan, 1998; Allen & Eckols (1997), Wu &
Yuan (2003).
Keller & Burkman, 1993, Dyson (2004),
Shneiderman & Plaisant, 2005.
Keller & Burkman, 1993; McGovern, Norton
& O'Dowd, 2002, Dyson (2004).
Allen & Eckols (1997)

Allen & Eckols (1997); McGovern, Norton &
O'Dowd, 2002.
Allen & Eckols (1997)


Winne (1995); Allen & Eckols (1997), Naidu
& Jarvela (2005).
Pettersson, 1999, Keller & Burkman, 1993;
Allen & Eckols (1997); McGovern, Norton &
O'Dowd, 2002, MacGregor & Lou (2004).
Keller & Burkman, 1993, Lee & Boling,
1999.
Keller & Burkman, 1993, MacGregor & Lou
(2004).
Weitl, SuIA, Kammerl, & Freitag, 2002

Keller & Burkman, 1993; McGovern, Norton
& O'Dowd, 2002., Lee & Boling, 1999,
Shneiderman & Plaisant, 2005.
Keller & Burkman, 1993; McGovern, Norton
& O'Dowd, 2002., Lee & Boling, 1999,
Shneiderman & Plaisant, 2005.
Keller & Burkman, 1993, Shneiderman &
Plaisant, 2005..
Pettersson, 1999, Lee & Boling, 1999,
Shneiderman & Plaisant, 2005.
Pettersson, 1985; McFarland, 1995, Lee &
Boling, 1999.


123






124


Graphic Elements


Item


Reference


Loss of overview (scrolling)


Visuals & graphics gain
attention
Favorable first impression

Picture depiction is accurate
Picture is aesthetically
satisfactory
Picture is technically
acceptable
Picture fits into the given
area
Picture fits in with the other
pictures on the page
Picture overload

Procurement (download)
time
Graphics used when
messages are complex
Graphics (charts) easy to
read
Meanings of icons
immediately obvious
Graphics support and
augment textual content

Captions placed nearer,
rather than farther, to
corresponding graphics
Consistent color scheme
Information w/color also
w/out color


Weitl, SuIA, Kammerl, & Freitag, 2002,
Dyson (2004), Shneiderman & Plaisant,
2005.
Fleming & Levie, 1993, Lee & Boling, 1999.

Keller and Burkman, 1993, Shneiderman &
Plaisant, 2005.
Pettersson 1998
Pettersson 1998; McGovern, Norton &
O'Dowd, 2002.
Pettersson 1998

Pettersson 1989

Pettersson 1989; McGovern, Norton &
O'Dowd, 2002.
Evans, Watson, and Willows,1987;
Pettersson, 1998
Pettersson 1998

March, 1983, McGovern, Norton & O'Dowd,
2002.
Pettersson, 1999, Keller & Burkman, 1993,
MacGregor & Lou (2004).
Pettersson, 1998

McFarland, 1995; Allen & Eckols (1997);
McGovern, Winne (1995); Norton &
O'Dowd, 2002.
Mayer, 1993, Robinson, Corlis, Bush, Bera
& Tomberlin (2003).

McFarland, 1995, Lee & Boling, 1999.
Pettersson, 1998









Realistic colors used in
graphics
Color media preferable to
black-and-white


Keller & Burkman, 1993; Allen & Eckols
(1997)
Simonson 1984; Allen & Eckols (1997)


125
















APPENDIX B
DELPHI PANEL WEB PAGE ELEMENTS IMPORTANCE CHECKLIST


Not Somewhat Very Critically Don't
Important Important Important Important Know
Text Elements
01 02 04 05 a
1 Whitespace (appropriate
use)
2 Text in visually distinct 2 4
01 02 04 05 O
segments
3 Short blocks of text 0 1 0 2 O 4 O 5 a
4 Running text in lowercase 2 4
01 02 04 05 O
letters
5 Lines of text in 8 to 10 4
01 02 04 05 O
words
6 Macrosignals (underline, 2 4 5
italics, bold)
7 Text color/background 2 L 4 5
contrast & brightness
8 Headings attract attention L0 1 0 2 a 4 O 5 a
9 Headings identify the 2 4
01 02 04 05 O
subject matter
10 Highlighted key words 0 1 0 2 O 4 a 5 a
11 Embedded hotlinks 0 1 0 2 4 a 5 a
12 Leading / Interlinear
01 02 04 05 O
spacing (appropriate use)
13 Loss of overview 5
01 02 04 05 O
(scrolling)
14 Column width 0 1 0 2 0 4 s 5 a
15 Abbreviations, acronyms
and initialisms 1 0 2 0 4 s 5
(appropriate use)
16 Avoid jargon, slang and 2 4
other unfamiar words 5
other unfamiliar words


126






127


Not Somewhat Very Critically Don't
Important Important Important Important Know
Graphic Elements

17 Provide metaphors or
quotes to make the L 1 0 2 O 4 1 5 L
unfamiliar familiar
18 Same information is
delivered in different 0 1 0 2 a 4 O 5 a
formats
19 Use of readable language 0 1 0 2 O 4 a 5 a
20 Well organized text 0 1 0 2 a 4 a 5 a
21 Explicit text 0 1 0 2 0 4 s 5 a
22 Content generalized to 2 4
01 02 04 05 O
specific
23 Familiar font 0 1 0 2 0 4 1 5 a
24 Legible font (san-serif) 0 1 0 2 0 4 s 5 a
25 Font size appropriate 0 1 0 2 0 4 s 5 a
26 Multiple type styles 2 4 L 5
(appropriate use)
27 Captions support graphics 0 1 0 2 0 4 1 5 a
28 Loss of overview 2 4
01 02 04 05 O
(scrolling)
29 Visuals & graphics gain 2 4 l 5
attention
30 Favorable first impression 0 1 0 2 0 4 1 5 a
31 Picture depiction is 2 4
01 02 04 05 O
accurate
32 Picture is aesthetically 2 L 4 5
satisfactory
33 Picture is technically 2 I 4 5 5
acceptable
34 Picture fits into the given 2 L 4 5
area
35 Picture fits in with the 2 4
01 02 04 05 O
other pictures on the page
36 Picture overload 0 1 0 2 0 4 s 5 a
37 Procurement (download) 1 2 4 l 5
time
38 Graphics used when 4
messages are complex
messages are complex






128


Not Somewhat Very Critically Don't
Important Important Important Important Know
Graphic Elements

39 Graphics (charts) easy to
01 02 04 05 O
read
40 Meanings of icons
01 02 04 05 O
immediately obvious
41 Graphics support and
01 02 04 05 O
augment textual content
42 Captions placed nearer,
rather than farther, to OL 1 0 2 O 4 O 5 a
corresponding graphics
43 Consistent color scheme OL 1 0 2 a 4 O 5 a
44 Information w/color also
01 02 04 05 O
w/out color
45 Realistic colors used in 2 4
01 02 04 05 O
graphics
46 Color media preferable to2 4
1 2 4 black-and-white5
black-and-white






129


Explanation of Checklist Items

Text Elements

Loss of overview: main message is presented "before the fold" (w/ out
scrolling)
Whitespace: used to set page elements apart
Text in visually distinct segments: sections or paragraphs
Short blocks of text: text is in 4-6 line blocks
Running text in lowercase letters: no "all caps"
Lines of text in 8 to 10 words
Macrosignals: use of underline, italics, bold for emphasis
Text color/background contrast & brightness
Typographic queuing:
Headings attract attention: headings are larger, bolder, or more colorful
than accompanying text
Headings identify the subject matter
Highlighted key words: highlighting for emphasis
Leading / Interlinear spacing: appropriate space between lines
Column width
Abbreviations, acronyms and initialisms: meanings should be identified
Avoid jargon, slang and other unfamiliar words
Provide metaphors or quotes to make the unfamiliar familiar
Same information is delivered in different formats: for example, directions
for driving by car are accompanied by a map
Use of readable language: appropriate grade level
Well organized text : in sections, paragraphs, w/ headings, etc.
Explicit text: message stated plainly and simply
Content generalized to specific
Familiar font
Legible font: san serif font is used
Font size appropriate: 12 pt font for running text
Multiple type styles: appropriate use (no more than 3 per page)
Captions support graphics


Graphic Elements

Loss of overview: can view without scrolling
Visuals & graphics gain attention
Favorable first impression
Picture depiction is accurate
Picture is aesthetically satisfactory
Picture is technically acceptable: not blurry, etc.
Picture fits into the given area (without overlap)
Picture fits in with the other pictures on the page






130


Picture overload: too many pictures
Procurement (download) time
Graphics used when messages are complex
Graphics (charts) easy to read
Meanings of icons immediately obvious
Graphics support and augment textual content
Captions placed nearer, rather than farther, to corresponding graphics
Consistent color scheme
Information w/color also w/out color: info that relies on color also
explained in text
Realistic colors used in graphics
Color media preferable to black-and-white: Hindenburg black and white
picture is OK; Exxon Valdez color picture is preferable















REFERENCES


Allen, B. S., & Eckols, S. L. (Eds). (1997). Handbook of usability principles.
Center for Learning, Instruction, & Performance Technologies, San Diego
State University, San Diego, CA.

Alley, L. R. (1996). An instructional epiphany. Information Technology, 28(2),
49-54.

American Vocational Association, Inc. (2000). Now Showing: The Virtual College
Tour. Techniques, 75(1), 13.

Assael, H. (1984). Consumer behavior and marketing action. Boston: Kent.

Babbie, E. (2001). The practice of social research (9th ed.). Belmont, CA:
Wadsworth.

Baines, A. (1998). Technology and tourism. Work Study, 47(5), 160-163.

Balakrishnan, R. (1991). Information acquisition and resource allocation
decisions. The Accounting Review, 66(1), 120-139.

Baloglou, S., & McCleary, K. W. (1999). A model of destination image formation.
Annals of Tourism Research. 26, 868-897.

Barrett, P. (2001, March). Assessing the reliability of rating data-revised.
Retrieved July, 2006, from http://www.pbmetrix.com/

Benson, P. J. (1985). Writing visually: Design considerations in technical
publications. Technical Communications Journal, 32, 207-216.

Berge, Z. L. (2000). Components of the online classroom. In R.E. Weiss, D.S.
Knowlton & B.W. Speck (Eds.) Principles of Effective Teaching in the
Online Classroom, 23-28.

Bevan, N. (1998). Usability issues in web site design. Retrieved October 2,
2004, from: http://www.npl.co.uk/npl/sections/us

Bieger, T. & Laesser, C. (2004). Information sources for travel decisions: Toward
a source process model. Journal of Travel Research, 42, 357-371.






132


Business Wire. (2002, January 22). Measure the impact of web site
improvements with WebCriteria's enhanced site analyst. Retrieved
February 1, 2002, from http://www.businesswire.com/

Carlson, N. R., (1993). Psychology: The science of behavior (4th ed.). Needham
Heights, MA: Allyn & Bacon.

Chall, J. S., Bissex, G. L., Conrad, S. S., & Harris-Sharpies, S. (1996).
Qualitative assessment of text difficulty: A practical guide for teachers and
writers. Cambridge, MA: Brookline.

Chandler, P. & Sweller, J. (1991). Cognitive load theory and the format of
instruction. Cognition and Instruction, 8, 293-332.

Cho, Y., & Fesenmaier, D. R. (2000). A conceptual framework for evaluating
effects of a virtual tour. In D. R. Fesenmaier, S. Klein, & D. Buhalis (Eds.),
Information and Communication Technologies in Tourism 2000,
Proceedings of the International Conference in Barcelona, Spain, 314-
323.

Chou, C. (2002). Developing the e-Delphi system: A web-based forecasting tool
for educational research. British Journal of Educational Technology,
33(2), 233-236.

CIAC (1998). US Department of Energy Computer Incident Advisory Capability
Information Bulletin 1-034: Internet Cookies, March 12, 1998. Retrieved
May 30, 2006, from http://www.ciac.org/ciac/bulletins/i-034.shtml.

Cisotto, L., & Boscolo, P. (1995, August). Effects of paragraph subtitles and
underlining on written summaries. Paper presented at the 6th European
Conference for Research on Learning and Instruction, Nijmegen.

Cooper, G. (1998). Research into cognitive load theory and instructional design
at UNSW. School of Education Studies, the University of New South
Wales, Sidney, NSW 2052, Australia.

Costello, A.B. & Osborne, J.W. (2005). Best practices in exploratory factor
analysis: Four recommendations for getting the most from your analysis.
Practical Assessment, Research & Evaluation, 10(7), 1-9.

Crocker, L., & Algina, J. (1986). Introduction to classical and modern test theory.
Orlando, FL: Harcourt Brace Jovanovich.

Csikszentmihalyi, M. (1975). Beyond boredom and anxiety. San Francisco:
Jossey-Bass.

Dalkey, N. & Helmer, 0. (1963). An experimental application of the Delphi
method to the use of experts. Management Science, 9(3), 458-467.






133


Dana, J. D. Jr., (1994). Learning in an equilibrium search model. International
Economic Review, 35(3), 745-771.

Daugherty, M. & Funke, B. (1998). University faculty and student perceptions of
web-based instruction. Journal of Distance Education, 13(1), 31-46.

Destination Marketing Association International (DMAI, 2006). FAQs about
CVBs. Retrieved September 11, 2006, from:
http://www.destinationmarketing.org/page.asp?pid=105

Eby, D. W., Molnar, L. J., Cai, L. A. (1999). Content preferences for in-vehicle
tourist information systems: an emerging tourist information source.
Journal of Hospitality & Leisure Marketing, 6, 41-58.

Ewing-Taylor, J. (1999). Student attitudes towards web-based courses.
University of Nevada-Reno, Retrieved September 11, 2006, from:
http://wolfweb.unr.edu/homepage/jacque/research/student_attitudes. html

Farquhar, J. D., & Surry, D. W. (1995). Reducing impositions on working
memory through instructional strategies. Performance & Instruction, 34(8),
4-7.

Fleming, M. L. & Levie, W. H. (Eds.). (1993). Instructional message design (2nd
ed.). Englewood Cliffs, N.J: Education Technology Publications.

Fleming, M. & Levie, W. H. (1995). Instructional message design: Principles from
the behavioral and cognitive sciences. Englewood Cliffs, NJ: Educational
Technology Publications.

Franks, K. (1996). Attitudes of Alaskan distance education students toward
media and instruction. The American Journal of Distance Education,
10(3), 64-68.

Frew, A. J. (2000). Information and communications technology research in the
travel and tourism domain: Perspective and direction. Journal of Travel
Research, 39, 136-145.

Gagne, E. D. (1985). The cognitive psychology of school learning. Boston: Little,
Brown.

Gagne, R. M., Briggs, L. J., & Wagner, W. W. (1988). Principles of instructional
design (3rd ed.). New York: Holt Reinhart Winston.

Gagne, R. W. (1977). The conditions of learning (3rd ed.). New York: Holt
Reinhart Winston.

Gallarza, M. G., Saura, I. G., & Garcia, H. C. (2002). Destination image: Towards
a conceptual framework. Annals of Tourism, 29(1), 56-78.






134


Geirland, J. (1996). Go with the flow. Wired Archive, Retrieved February, 2002,
from: http://www.wired.com/wired/archive/4.09/czik_pr.html

Gibson, J. W. (2001). The journey to cyberspace: Reflections from three online
business professors. Advanced Management Journal, 66(1), 30-34.

Global Reach. (2004). Global Internet Statistics. Retrieved November, 2005,
from http://global-reach.biz/globstats/index.php3.

Goldfarb, I., & Kondratova, I. (2003, June). Using visual materials to engage
learners. Proceedings of the ED-Media 2003 ACEE World Conference on
Educational Media, Hypermedia, and Telecommunications. Honolulu,
Hawaii, USA.

Govers, R., Jansen-Verbeke, M., & Go, F. M. (2000). Virtual tourist destinations:
Assessing their communication effectiveness with and through foreign
intermediaries. In D. R. Fesenmaier, S. Klein, & D. Buhalis (Eds.),
Information and Communication Technologies in Tourism 2000,
Proceedings of the International Conference in Barcelona, Spain, 93-103.

Gretzel, U., Yuan, Y., & Fesenmaier, D. R. (2000). Preparing for the new
economy: Advertising strategies and change in destination marketing
organizations. Journal of Travel Research, 39, 146-156.

Guthrie, J. T. (1988). Locating information and documents: Examination of a
cognitive model. Reading Research Quarterly, 23, 178-199.

Hall, R. H. (2001). Web-based training site design principles: a literature review
and synthesis. (165-172). In B. Kahn (Ed.), Web-based training,
Englewood Cliffs, NJ: Educational Technology Publications.

Henney, M. (1981). The effect of all-capital print versus regular mixed print as
displayed on a micro-computer screen on reading and speed accuracy.
Education Communications and Technology Journal, 31, 126.

Hodkinson, C. & Kiel, G. (2003). Understanding web information search
behavior: An exploratory model. Journal of End User Computing, 15(4),
27-48.

Hubbard, A. (1998). What it means to teach online. Computers and
Geosciences, 24(7), 713-717.

Huizingh, E. (2000). The content and design of web sites: an empirical study.
Information and Management, 37, 123-134.

Hull, L. (2004). Accessibility: It's not just for disabilities any more. Interactions,
11(2), 36-41.






135


Innovations in Distance Education. (1998). An emerging set of guiding principles
and practices for the design and development of distance education. The
Pennsylvania State University. Retrieved February 1, 2002, from
http://www.outreach.psu.edu/de/ide/guiding_principles/

Innovations in Distance Education. (1998). An emerging set of guiding principles
and practices for the design and development of distance education. The
Pennsylvania State University. Retrieved February 1, 2002, from
http://www.outreach.psu.edu/de/ide/guiding_principles/

Jesitus, J. (1998). On-line CVB's lure delegates, book business. Hotel & Motel
Management, 213(6), 15, 67.

Jonassen, D. H. & Kirschener, P. A. (1982). Explicit techniques for structuring
text. In D. H. Jonassen (Ed.). The technology of text: Principles for
structuring, designing, and displaying text, Vol. 2. Englewood Cliffs, NJ:
Educational Technology Publications.

Jung, T. H., & Butler, R. (2000). The measurement of the marketing
effectiveness of the internet in the tourism and hospitality industry. In D.
R. Fesenmaier, S. Klein, & D. Buhalis (Eds.), Information and
Communication Technologies in Tourism 2000, Proceedings of the
International Conference in Barcelona, Spain, 460-472.

Kedia, B. L. & Bhagat, R. S. (1988). Cultural constraints on transfer of technology
across nations: Implications for research in international and comparative
management. Academy of Management Review, 13 (4), 559-571.

Keller, J. M. & Burkman, E. (1993). Motivation principles. In M. Fleming & W.H.
Levie (Eds.) Instructional message design: Principles from the behavioral
and cognitive sciences (pp. 3-53) (2nd ed.). Englewood Cliffs, NJ:
Educational Technology.

Keller, W. (2002). Geographic localization of international technology diffusion.
The American Economic Review, 92 (1), 120-142.

Kiiski, S. & Pohjola, M. (2002). Cross country diffusion of the internet. Information
Economics and Policy, 14, 297-310.

Knowlton, D. S. (2000). A theoretical framework for the online classroom: A
defense and delineation of a student-centered pedagogy. In R.E. Weiss,
D.S. Knowlton & B.W. Speck (Eds.) Principles of Effective Teaching in the
Online Classroom, 5-14.

Kormier, K. A. (1999). Netstorming: Web site development strategies. Strategic
Finance Magazine, 80(11), 64-67.






136


Kovacs, P., & Rowell, D. (2001). The merging of systems analysis and design
principles with web site development: One university's experience. T.H.E.
Journal, 28(6), 60-65.

Kroil, P., Kapsammer, E., Pr6ll, B., Retschitzegger, W., & Wagner, R. R. (1997).
Online booking on the net: problems and solutions. Research Institute for
Applied Knowledge Processing (FAW), University of Linz, Austria.
Retrieved September 11, 2006, from: ftp://ftp.ifs.uni-
linz.ac.at/pub/publications/1 997/1197.pdf

Laesser, C, & Beiger, T. (2004). Information sources for travel decisions: Toward
a source process model. Journal of Travel Research, 42(4), 357-371.

Lake, D. (2001) Americans go online for travel information. CNN, June 14, 2001.
Available: http://www.cnn.com/2001/TECH/internet/06/14/travelers.use.net
.idg/index.html

Leiner, B. M., Cerf, V. G., Clark, D. D., Kahn, R. E., Kleinrock, L., Lynch, D. C.,
Postel, J., Roberts, L. G., & Wolff, S. (2003). A brief history of the
Internet. Internet Society, retrieved November 2, 2005 from
http://www.isoc.org/internet/history/brief.shtml

Linacre, J. M. (2002). Judge ratings with forced agreement. Rasch Measurement
Transactions, 16(1), 857-858.

Linstone, H. A. & Turoff, M. (2002). The Delphi method: Techniques and
applications. Retrieved April 12, 2005, from:
http://www.is.njit.edu/pubs/delphibook/

Lohr, L. L. (2000). Designing the instructional interface. Computers in Human
Behavior, 16, 161-182.

MacDonald, L., & Caverly, D. C. (2001). Delivering study skills face-to-face at a
distance. Journal of Developmental Education, 24(3), 44-45.

MacLennan, R. N. (1993). Interrater reliability with SPSS for Windows 5.0 The
American Statistician, 47(4), 292-296.

Machlis, S. (1997). Profits elude travel sites. Computerworld, 32(1), 53-54.

Mayer, R. (1992). Cognition and instruction: Their historic meeting within
educational psychology. Journal of Educational Psychology, 84, 405-412.

Mayer, R. E. & Moreno, R. (2002). Aids to computer based multimedia learning.
Learning and Instruction, 12, 107-119.

McFarland, R. (1995). Ten design points for the human interface to instructional
multimedia. THE Journal, 22(7), 67-69.






137


McGovern, G., & Norton, R. (2002). Content critical. London, England: Pearson.

McGovern, G., Norton, R. & O'Dowd, C. (2002). The web content style guide.
London, England: Pearson.

McKenna, J., Avery, R., & Schuchardt, J. (2000). Technology strategies for
enhancing learning. Consumer Interests Annual, 46, 200-240.

Metros, S. E. (2001). Engaging online learners. Lifelong Learning in Europe,
6(2), 85-95.

Meyer, R. (1982). A descriptive model of consumer information search behavior.
Marketing Science, 1(1), 93 -121.

Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits
on our capacity for processing information. The Psychological Review, 63,
81-89.

Miller, M. J. (1996). Being virtual. PC Magazine, 15(3), 29.

Montilva, J. A., Sandia, B., & Barrios, J. (2002). Developing instructional web
sites: A software engineering approach. Education and Information
Technologies, 7(3), 201-224.

Moorthy, S., Ratchford, B., & Talukdar, D. (1997). Consumer information search
revisited: Theory and empirical analysis. Journal of Consumer Research,
23, 263-277.

Morrison, G. R. & Guenther, P. F. (2000). Designing instruction for learning in
electronic classrooms. In R.E. Weiss, D.S. Knowlton & B.W. Speck (Eds.)
Principles of Effective Teaching in the Online Classroom, 15-22.

Motley, L. B. (2000). The 5 elements of web site alchemy. Bank Marketing,
32(3), 42-43.

Murry, J. W., & Hammons, J. O. (1995). Delphi: A versatile methodology for
conducting qualitative research. The Review of Higher Education, 18(4),
422-436.

Naisbitt, J. (1984). Megatrends: Ten new directions transforming our lives. New
York: Warner.

Newman, F., & Scurry, J. (2001). Online technology pushes pedagogy to the
forefront. The Chronicle of Higher Education, 47(44), B7.

OneStat (2005). Retrieved September 11, 2006, from
httD://www.onestat.com/html/aboutus Dressbox31.html






138


Palmer, A. & McCole, P. (2000). The role of electronic commerce in creating
virtual tourism destination marketing organizations. International Journal
of Contemporary Hospitality Management, 12(3), 198-204.

Pan, B., & Fesenmaier, D. R. (2000). A typology of tourism related web sites: Its
theoretical background and implications. In D. R. Fesenmaier, S. Klein, &
D. Buhalis (Eds.), Information and Communication Technologies in
Tourism 2000, Proceedings of the International Conference in Barcelona,
Spain, 381-396.

Petterson, R. (1999, July). Attention: An information design perspective. Paper
presented at the International Institute for Information Design (IIID),
Vienna, Austria.

Pettersson, R. (1998, October). Image functions in information design. Paper
presented at the 30th International Conference of the International Visual
Literacy Association, The University of Georgia, Athens, GA.

Pike, S. (2002). Destination image analysis: A review of 142 papers from 1973 to
2000. Tourism Management, 23(5), 541-549

Poling, C. (1999). Snaring customers on the web: You'd better learn how-and
fast. New Mexico Business Journal, 23(5), 42-43.

Poulton, E. C., & Brown, C. H. (1968). Rate of comprehension of an existing
teleprinter output and of possible alternatives. Journal of Applied
Psychology, 52, 16-21.

Rankin, R. O. (1989). The development of an illustration design model.
Educational Technology Research and Development, 37(2), 25-46.

Raward, R. (2001). Academic library web site design principles: Development of
a checklist. Australian Academic & Research Libraries, 32(2), 123-136.

Rayman-Bacchus, L., & Molina, A. (2001). Internet-based tourism services:
business issues and trends. Futures, 33(7), 589-591.

Reigeluth, C. M. (Ed). (1983). Instructional design theories and models: An
overview of their current status. London: Erlbaum.

Rosen, D. E., & Purinton, E. (2002). Web site design: Viewing the web as a
cognitive landscape. Journal of Business Research, 57, 787-794.

Schley, R. (1997). Travel planning online: Virtual visits give travelers a head
start. The Futurist, 31(6), 12.

Schuman, J. (1996). Don't tangle up your online visitors with web site design.
Hudson Valley Business Journal, 7(15), 13.






139


Seal, K. C., & Przasnyski, Z. H. (2001). Using the world wide web for teaching
improvement. Computers and Education, 36, 33-40.

Sherman, L. W. (2000). Postmodern constructivist pedagogy for teaching and
learning cooperatively on the web. Cyberpsychology and Behavior, 3(1),
51-57.

Sigala, M., Airey, D., Jones, P., & Lockwood, A. (2000). The diffusion and
application of multimedia technologies in the tourism and hospitality
industries. In D. R. Fesenmaier, S. Klein, & D. Buhalis (Eds.), Information
and Communication Technologies in Tourism 2000, Proceedings of the
International Conference in Barcelona, Spain, 397-407.

Simonson, M. R. (1984). Media and persuasive messages. Instructional
Innovator, 29(2), 23-24, 48.

Snepenger, D., Meged, K., Snelling, & Worral, K. (1990). Information search
strategies by destination naive tourists. Journal of Travel Research,
29(1), 13-16.

Spiro, R. J., Feltovich, P. J., Jacobson, M. I. & Coulson, R. L. (2002). Cognitive
flexibility, constructivism, and hypertext: Random access instruction for
advanced knowledge acquisition in ill-structured domains. Retrieved
September 11, 2006, from:
http://phoenix.sce.fct.unl.pt/simposio/Rand_Spiro.htm

StatSoft, Inc. (2006). Electronic statistics textbook. Tulsa, OK: StatSoft.
Available: http://www.statsoft.com/textbook/stathome.html.

Stemler, S. E. (2004). A comparison of consensus, consistency, and
measurement approaches to estimating interrater reliability. Practical
Assessment, Research & Evaluation, 9(4). Retrieved July 25, 2006 from
http://PAREonline.net/getvn.asp

Sweller, J. & Chandler, P. (1991). Evidence for cognitive load theory. Cognition
and Instruction, 8(4), 351-362.

Sweller, J. (1994) Cognitive load theory, learning difficulty, and instructional
design. Learning and Instruction, 4, 295-312.

Sweller, J., Van Merrienboer, J.J.G. & Paas, F.W.C. (1998) Cognitive
architecture and instructional design. Educational Psychology Review 10,
251-296.

Tabbers, H. K. (1999). The effect of presentation mode on cognitive load:
Teaching cognitive skills with text and pictures. Open University of the
Netherlands, Heerlen.






140


Teh, G. P. L., (1999). Assessing student perceptions of internet-based online
learning environments. International Journal of Instructional Media, 26,
397-402.

Thompson, G. (1990). How can correspondence-based distance education be
improved? Journal of Distance Education, 15(1), 61-70.

Tierney, P. (2000). Internet-based evaluation of tourism web site effectiveness:
Methodological issues and survey results. Journal of Travel Research,
39, 212-219.

Tinker, M. A. (1963). The legibility of print. Ames, IA: State University Press.

Tsai, P., Cambiano, R.L., De Vore, J.B., & Harvey, R.L. (2000). Graduate
students perceptions of distance learning. Academic Exchange, Fall, 62-
66.

Turoff, M., Hiltz, S. R., Li, Z., Wang, Y., Cho, H., Yao, X., (2004). Online
collaborative learning enhancement through the Delphi method.
Proceedings of the OZCHI 2004 Conference, November 22-24, University
of Wollongong, Australia. Retrieved October, 2005, from:
http://web.njit.edu/-turoff/Papers/ozchi2004.htm

Urbina, S. P. (1994). Attention. In R. J. Corsini, Encyclopedia of Psychology (2nd
ed.). New York: John Wiley & Sons.

Vogt, C. A., & Pennington-Gray, L. (2002, June). Using behavioral segmentation
to understand the influence of information on welcome center stoppers.
Paper presented at the annual meeting of the Travel and Tourism
Research Association, Arlington, VA.

Vogt, C. A., & Fesenmaier, D. R. (1998). Expanding the functional information
search model. Annals of Tourism Research, 25(3), 551-578.

Weber, K., & Roehl, W. S. (2002). Profiling people searching for and purchasing
travel products on the World Wide Web. Journal of Travel Research,
37(3), 291-298.

Weitl, F., Sui, C., Kammerl, R., & Freitag, B. (2002). Presenting complex e-
learning content on the web: A didactical reference model. Proceedings of
World Conference on E-Learning in Corporate, Government, Healthcare
and Higher Education, Montreal, Canada.

Weston, C., Gandell, T., McAlpine, L., & Finkelstein, A. (1999). Designing
instruction for the context of online learning. The Internet and Higher
Education, 2(1), 35-44.









Whalen, S. P. (1997). Assessing flow experiences in highly able adolescent
learners. (ERIC Document Reproduction Service No. ED 409 382)

Wilhelm, W. J. (2001). Alchemy of the oracle: The delphi technique. The Delta
Phi Epsilon Journal, 43(1), 6-26.

Winne, P. (1995, August). Challenges to the development of self-regulated
learning. Paper presented at the 6th European Conference for Research
on Learning and Instruction. Nijmegen.

Witt, G. A. (1981). How to present information that viewers will remember.
Instructional Innovator, 26(9), 37-43.

Yacci, M. (2000). Interactivity demystified: A structural definition for distance
education computer based instruction. Educational Technology 40(4), 5-
16.















BIOGRAPHICAL SKETCH

Charles W. Lane was born in Brooklyn, New York, and grew up in Miami,

Florida. He graduated from the University of Florida with a Bachelor of Science

degree in botany in 1994. He studied micropropagation of wetland plant species

with Dr. Michael Kane and graduated with a Master of Science in environmental

horticulture from the University of Florida in 1999. His master's thesis was

entitled Tissue Culture Propagation of Sagittaria latifolia. After graduation with a

Doctor of Philosophy degree in tourism, recreation and sport management, he

intends to pursue a career in higher education.


142





PAGE 1

DESTINATION MESSAGE DESIGN ON TRAVEL AND TOURISM INFORMATION WEB SITES By CHARLES W. LANE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PH ILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

Copyright 200 7 by Charles W. Lane

PAGE 3

iii ACKNOWLEDGMENTS This study was conducted with the assistance and support of numerous individuals. My sincerest gratitude is extended to all those who contributed. My si ncere appreciation goes to Dr. Stephen Holland, my committee chair who ha s provided me with so much support. My heartfelt thanks also go es to Dr. Heather Gibson, who provided thoughtful and relevant feedback and whose consta nt encouragement has kept me going. Thanks to Dr. Lori Pennington Gray for adding so much to my doctoral program. The expertise of m y external committee member, Dr. Colleen Swain, has been invaluable throughout my program. I am deeply indebted to Dr. Bri jesh Thapa. Without his support this project could not have been completed. I would also like to express my thanks to my fellow doctoral students, especially Jung Eun Kim, Louisa Meyer, and Lisa Pen n isi, who have made this experience enjoyable and provide d much needed support Valuable input also came from Dr. Benjamin Lok and Dr. Lee Mullally The hard work of Michell York and Mario Klemmer contributed significantly to this study and their efforts are deeply appreciated S pecial thanks are also due to my masters guru Dr. Michael Kane without whom my graduate education would not have been possible. Finally, and most importantly, the love, support, and encouragement that I have received from my family have made this endeavo r not only possible, but also worthwhile.

PAGE 4

iv TABLE OF CONTENTS ACKNOWLEDGMENTS ................................ ................................ .................... iii LIST OF TABLES ................................ ................................ .............................. v i LIST OF FIGURES ................................ ................................ ........................... vi i i ABSTRACT ................................ ................................ ................................ ....... ix CHAPTER 1 INTRODUCTION ................................ ................................ ..................... 1 Tourism and the Web ................................ ................................ .............. 4 Instructional Message Design ................................ ................................ 11 Theoretical Framework ................................ ................................ .......... 12 Purpose of the Study ................................ ................................ ............. 13 2 REVIEW OF THE LITERATURE ................................ ........................... 15 The World Wide Web ................................ ................................ ............. 15 Technology in Tourism ................................ ................................ .......... 23 Internet Web Site Evaluation ................................ ................................ 31 Instructional Message Design ................................ ................................ 33 Cognitive Load Theory ................................ ................................ ........... 50 Application of Cognitive Load Theory and Instructional Mes sage Design Principles to Tourism Information Search ............................. 54 3 METHODS ................................ ................................ ............................. 58 Methods: Research Questions 1 4 ................................ ........................ 59 Methods: Research Question 5 ................................ ............................. 68 4 RESULTS ................................ ................................ .............................. 72 Results of Data Analysis........................................................................78 Summary...............................................................................................96

PAGE 5

v 5 DISCUSSION ................................ ................................ ........................ 98 Discussion of the Results ................................ ................................ ....... 98 Limi tations and Delimitations of the Study ................................ ........... 116 Implications for Practice ................................ ................................ ....... 117 Implications for Future Research ................................ ......................... 119 APPENDIX A INSTRUCTIONAL MESSAGE DESIGN ELEMENTS ASSESSED ON CVB WEB SITES ................................ ................................ ..... 122 B DELPHI PANEL WEB PAGE ELEMENTS IMPORTANCE CH ECKLIST ................................ ................................ ................... 126 REFERENCES ................................ ................................ ............................... 13 1 BIOGRAPHICAL SKETCH ................................ ................................ ............. 1 42

PAGE 6

vi LIST OF TABLES Table page 2 1 Support in the literature for web site content elements .................... 20 2 2 Web site evolution by generation ................................ ..................... 34 3 1 Sample from CVB world regions ................................ ...................... 60 3 2 Rater summary scores on primary component ................................ 65 4 1 Response variability of the Likert type items ................................ .... 73 4 2 Response location of the survey items ................................ ............ 74 4-3 Item discrimination of Likert items (alpha=.862)...............................7 6 4-4 Item analysis composite...................................................................77 4 5 Prevalence of instructional message design elements on CVB web sites ................................ ................................ ..................... 79 4 6 Ratings of message design element use on CVB web sites ............ 81 4 7 Factor analysis and reliability of the message design variables ....... 8 5 4 8 Message desi gn elements exhibiting significant differences between regions ................................ ................................ .......... 8 7 4 9 Graphic presentation significant difference by region ...................... 88 4 1 0 Broadband download times by world region ................................ .... 89 4 1 1 Flesch Kincaid grade level readability scores by regio n .................. 8 9 4 1 2 Legibility of text by region ................................ ................................ 90 4 1 3 Color vision impairment mean scores for text by region .................. 9 0 4 1 4 Legibility of headings by region ................................ ........................ 91 4 1 5 Significant difference between text and heading color difference .... 9 1 4 1 6 Significant difference between t ext and heading brightness difference ................................ ................................ .................... 9 1 4 1 7 Color vision impairment mean scores for headings by region .......... 92

PAGE 7

vii 4 18 Text / headings difference: Protanopia ................................ ............ 9 2 4 1 9 Text / headings difference: Deuteranopia ................................ ........ 93 4 2 0 Text / headings difference: Tri tanopia ................................ .............. 93 4 2 1 Delphi panel importance ratings of message design elements ........ 9 4

PAGE 8

viii LIST OF FIGURES Figure page 1 1 Assaels (1984) model of consumer information acquisition and processing ................................ ................................ ..................... 8 1 2 The relationship supportive message design, cognitive lo ad, and information processing ................................ ................................ 13 2 1 The role of supportive message design in the facilitation of consumer information acquisition and processing ...................... 55 4 1 Scree plot of factor loadings ................................ ............................. 8 3

PAGE 9

ix Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DESTINATION MESSAGE DESIGN ON TRAVEL AND TOURISM INFORMATION WEB SITES By Charles W. Lane August 20 07 Chair: Stephen Holland Major: Health and Human Performance The rapid expansion of the Internet has changed the way people communicate. The Internet is the primary source of informati on in many aspects of our lives, including travel planning. Tourists expect to be ab le to gather online the information they need to plan their travel. Convention and Visitors Bureaus (CVBs) exist to provide travelers with information about destina tions. Unlike the sites of airlines, hotels, and attractions, the primary mission of most CVB web sites is not sales. Instead a CVB web site inform s users about the location it represents. Literature from instructional message design and learning theory research support s the notion that an effective informational web site, such as that of a CVB, should employ supportive message design elements to reduce the cognitive load of the user to enhance information acquisition and processing.

PAGE 10

x Evaluating web sites according to instructional message design principles is a method of appraising web site effectiveness in enhancing information acquisition and processing ; however, in the commercia l sector, such techniques have not typically been applied The purp ose of this study was to examine the use of supportive message design elements on CVB web sites worldwide. A content analysis of 588 sites was conducted to determine which elements w ere used and to what extent. Statistical analyses were conducted to dete rmine if there were differences among regions of the world in the use of effective design principles. Results indicated that many message design practices were nearly ubiquitous on CV B web sites. Web sites were more effective in their use of visual display elements than in their use of content presentation elements. Message design elements can be reliably grouped into four dimensions: graphics presentation, text presentation, graphi cs/text support, and visual text. Regional differences were found on nine message design elements and one underlying factor. Expert web developers considered functionality and aesthetics to be the most essential features, revealing a lack o f understanding in the field of the importance of the design of message content delivery R esults are discussed in the context of the theoretical framework and previous research, and r ecommendations for practice and for future research are provided.

PAGE 11

CHAPTER 1 INTRODUCTION Effective communication is critical in modern society. We are living in what is widely known as the Information Age a time in which information and the manner in which information is communicated is more importa nt than ever before. Increasingly, communication of information occurs via computer technology. In fact, according to Naisbitt (1984), computer technology is to the information age what mechanization was to the industrial revolution (p. 28). One of t h e most pervasive use s of computers is to access the Internet. In its relatively short lifespan, the World Wide Web has grown to a massive size. The Internet is a borderless system of computer networks that allo ws any computer to communicate with any or all other computers on the network. It is, simultaneously, a medium for human computer interaction, a vehicle for information dissemination, and a forum for the sale of goods and services. Although these functions all occur without respect to geographic placement, the diffusion of the Internet itself has been characterized by sharp disparities in technological development and accessibility (Kiiski & Pohjola, 2002). The Internet had its genesis in the United States as a tool used largely by scientific researche rs (Rosen & Purinton, 2002). Even at its inception, the Internet was a cluster of related technologies where accessibility depended on

PAGE 12

2 2 both hardware and infrastructure: a computer, a remote host, and connective telephone lines (Kiiski & Pohjola, 2002). T he global diffusion of the Internet has been achieved relatively quickly, but not with the uniformity that is commonly believed. The primary factors that influence the transfer of technologies across nations are the characteristics of the technology, cultu ral variation, societal differences, and the absorptive capacity of the recipient region. As a result, the process of technological transfer is not linear ( Kedia & Bhagat, 1988) The degree of availability of Internet hardware and infrastructure has been influenced by the ambivalence of governments and the corresponding irregularities in government policies (Goodman, Burkhart, Foster, Press, Tan, & Wo o dard, 1998). For developed countries, the diffusion of the I nternet and related technologies has been su pported by governments and large corporations through the creation of research and academic networks that serve as the backbone of the enterprise. For most other countries, these types of national technological backbones are beyond the scope of their finan cial resources, and, consequently, their national priorities ( Goodman, Press, Ruth, & Rutkowski, 1994). By September 2004, there were an estimated 287.5 million English language web users online and an additional 516.7 million non English language web user s (Global Reach, 2004). Current statistics estimating the total number of web pages on the Internet have generally been offered alongside a warning of their unreliability. The majority of homes in America now have at least one computer, and 64% of all Ame ricans have used the Internet in the past year

PAGE 13

3 3 (Rosen & Purinton, 2002). The Internet has become the primary technological medium for delivering online content, such as the information, features, or services found on web sites (Huizingh, 2000), in addition to instructional material (Montilva, Sandia, & Barrios, 2002). Internet connectivity is strongly correlated to GDP per capita across world regions, where people in wealthy, developed areas enjoy a higher rate of access than those in undeveloped areas. S ince information technology is thought to be a driver of economic growth, the inequalities in the diffusion of I nternet technology may be a contributing factor to the disparity of income potentials between the richer and poorer world regions (Kiiski & Poh jola, 2002). In addition, technology can contribute to sustained inequality due to its uneven distribution throughout areas of the world. Given the importance of the Internet, the level of diffusion can influence whether an area can sustain its place in the global economy (Harhittai, 1999), and the potential for impact on the tourism industry is great (Frew, 2000) Sigala, Airey, Jones, and Lockwood (2000) explained however, that due to the nature of the tourism industry, characterized by multiple small businesses, the use of technology by tourism professionals has been slow to diffuse to some areas. There is widespread agreement that the Internet is evolving as a dynamic electronic marketplace at an unprecedented pace ( Rayman Bacchus & Molina, 2001). Bu sinesses, then, need to orient their marketing strategies to address wider diffusion of the potential consumer base (Rai, Ravichandran, & Samaddar, 1998). Over the past few years, the Internet has undergone a wave of

PAGE 14

4 4 commercialization where attention has been focused on the use of the global information infrastructure to support commodity and other commercial services (Leiner et al., 2003). Projections of continued growth serve to inform providers of I nternet products and services of the potential worldw ide market. Because the Internet is a phenomenon that grew very quickly, the rush to generate a web presence prevented many web content providers from researching the needs and wants of web consumers (Hodkinson & Kiel, 2003) prior to going online. Conseq uently, web page designers, where design refers to the way that content is made available to web site visitors (Huizingh, 2000), produced a variety of information delivery formats in a research vacuum. Belatedly, content providers have turned to empirical research to examine the needs and wants of web based consumers (Hodkinson & Kiel, 2003). Tourism and the Web As a correlate to the growth of the Internet, there is evidence that the classic model of the tourism product is changing from mass tourism m arkets dominated by a few providers to a broader array of more specialized types of tourism providers (Sigala et al. 2000). This has, in part, been shaped by the evolution of technologies important to the development of the burge oning travel industry. The arrival of the jumbo jet in the early 1970s, along with computer based reservations and travel information systems, offered travelers greater destination choice, but was still dependent on the services of various travel agencies to enable that choice.

PAGE 15

5 5 The nature of the tourism industry, sometimes characterized by multiple small businesses, has resulted in the slow diffusion of technology and Internet access in some areas (Sigala et al., 2000). For the tourism industry, the Int ernet has also emerged as an electronic travel agent an easy access platform that brings travel information to consumers directly without using traditional intermediaries like travel agents (Kroi Kapsammer, Prll, Retschitzegger, & Wagner, 1997). The dev elopment of the Internet and World Wide Web has provided individual travelers with a much greater degree of autonomy in travel information search and destination selection. At the same time, the Web has enabled small to medium sized tourism provid ers to compete with mass tourism destinations for individual travelers attention ( Rayman Bacchus & Molina, 2001). It is widely accepted that the travel industry is more likely to be affected by technological advances than other industries, as trave l has consistently placed among the top three product categories purchased via the Internet (Weber & Roehl, 1999). Machlis (1997) reported that the largest revenue generator among consumers on the Internet was the travel industry. The traveler who was onc e forced to rely on a travel agent can now make direct inquiries of airlines, hotels, and destinations (Baines, 1998). Destination marketing organizations such as state and national tourism offices and convention and visitor bureaus (CVBs), can pr ovide a web presence for many smaller tourism businesses. Convention and visitors bureaus, in particular, seldom have a product of their own to represent, but instead act as

PAGE 16

6 6 information brokers between tourism destinations and potential visitors (P almer & McCole, 2000). That information brokerage is enhanced by the maturation of the Internet and World Wide Web where delivery of information through high quality message design is possible. According to the Destination Marketing Association International (DMAI, 2006), the purpose of a CVB is to represent selected destinations to meeting professionals, tour operators, and individual visitors. The mission of a local CVB is to present information about a geographical area (e.g., city, state, region) to the public, with the goal of increasing visitation to the area through tourism and meetings. Publicly funded bureaus, like some CVBs, have had to become increasingly transparent with respect to expenditures of public dollar s. The shrinking of the tax base in some areas along with erosion in city and county budgets makes a target of the revenue generated from hotel room bed taxes (Schweitzer, 1997). Some bureaus are facing budget cuts despite their classification as revenue producers. As a result, CVBs have responded by initiating service improvements targeted at remaining competitive in the online marketplace. Adapting their traditional ways of doing business to the Internet has become the focus of CVBs nationwide. Schweit zer ( 1997) explained that CVBs now attract new business by dedicating resources to using technology as an effective way to promote the destination A nationwide trend among CVBs is enhancement of web based services to provide more comp rehensive information

PAGE 17

7 7 to various visitors and yield both advantages for consumers and benefits for the bureaus themselves Providing information to potential visitors is one of the primary goals of an organization's web site ( Jesitus, 1998) and central to the primary mission of a CVB. Since the Internet has become an integral part of daily life in the US and elsewhere, it has developed importance to CVBs. Lake ( 2001 ) found that 93% of travelers used the web to gather t ravel related information To reach web savvy travelers, maximizing the effectiveness of their web presence should be a priority for CVBs. The primary external source for travel and tourism information is, increasingly, a destinations web si te. Travel consumers have readily accepted this technology because it is convenient and easy to work with (Kroi et al., 1997). Vogt and Fesenmaier (1998) considered information search behaviors in the travel and tourism industry, and expanded the role of tourism information search from a primarily marketing purview into a broader communication context. According to Vogt and Fesenmaier, it is well established that visitors to a web site often come primarily to gather information. Tourism information se ekers primarily collect and use information for functional reasons, such as planning or taking trips. A later decision to purchase may be greatly influenced by the information gathering phase of the visit. To guide potential tourists toward a purchase dec ision, tourist information should appeal to a variety of search needs to first capture the attention of potential travelers. Using Assaels (1984) model of consumer information acquisition and processing (Figure 1 1), Vogt and

PAGE 18

8 8 Fesenmaier explained the trav elers steps in making a tourism purchasing decision. Figure 1 1. Assaels (1984) model of consumer information acquisition and processing. Evaluation of Tourism Web Site Effectiveness The technological and financial barriers to establishing a busines s web presence on the Internet are few. As a result, tourism professionals have established a web presence that encompasses diverse expectations and formulas for success ( Rayman Bacchus & Molina, 2001). Unfortunately, the ease of authoring web pages has c reated quality control issues. The ubiquitous nature of web site creation provides a widespread opportunity for inexperienced information providers to create unending information delivery systems that are difficult to use (Bevan, 1998). Web pages are oft en not subject to the same quality criteria as can be found in more traditional forms of publishing (McGovern

PAGE 19

9 9 & Norton, 2002). Leading travel and tourism sites, however, are visually attractive, provide detailed and current content, and succeed at maintai ning the interest of their site visitors. The success of leading tourism sites suggests a competitive advantage for web sites that are visually appealing and information rich ( Rayman Bacchus & Molina, 2001). Central to the principles of instructional mes sage design is the presentation of information in a way that is visually appealing and that will attract and maintain the users attention ( Flem ing & Lev ie 1993 ). Evaluation of web site effectiveness is necessary due to the volume of re tail sales generated from Internet commerce $74 billion in 2002 (Rosen & Purinton, 2002) and due to the substantial costs associated with establishment and maintenance of web sites (Tierney, 2000). The complexity of the web itself, however, poses problems for creation of standardized measures of web site evaluation (Gretzel, Yuan, & Fesenmaier, 2000). Tabulating direct sales from a site is one method of evaluation, but many tourism web sites, such as those of CVBs, do not sell directly to the consumer and cannot be measured in this manner. According to Tierney (2000), other common sources of data for web site evaluation are number of hits, number of user sessions, and online feedback from site users incorporating interactivity into a web s ite can enable the visitor to relay a great deal of personal information back to the host Another tracking strategy employed by commercial web sites is the use of cookies to monitor repeat visits, viewer preferences, and page viewings

PAGE 20

10 10 Statement of the Problem Access to web information can be influenced by a variety of technical issues, such as the speed of the Internet service provider (ISP), the settings and capabilities o f the computer involved, and the web browser being used. A combination of these features defines the technical environment that affects the display of information and the web site features that will ultimately be available to the information searcher (Hod kinson & Kiel, 2003). Although web site developers cannot control the technical environment of the end user, successful web site development remains, in part, contingent upon the developers awareness of graphic design and web usability (Bevan, 1998). If a web site does not meet the expectations and needs of its visitors, then it will not meet the needs of the organization represented on the site. Developing an informational web site necessitates the inclusion of instructional, structural, functional, and aesthetic properties (Montilva, Sandia & Barrios, 2002). When creating web sites with the primary purpose of providing information, web developers may use general purpose methods similar to those used for creating commercial web pages. The problem with t hat practice is that it does not address those requirements that are intrinsic to informational web sites where retention is the desired outcome. To date, the discipline of instructional message design has not been applied to the evaluation of th e travel and tourism web presence as a learning experience, as most web site developers do not consider those aspects that are specific to the information acquisition process (Montilva, Sandia, & Barrios,

PAGE 21

11 11 2002). Given that a primary function of CVBs is to present information to potential visitors, travel industry professionals could benefit from understanding ways to convey that information more effectively. Instructional Message Design Digital technology and interactive multimedia have had a significant i mpact in the field of education, where web based e learning environments utilize the power of the Internet to create a context where retention of in formation is supported and enabled. Winne (1995) suggested that learning is affected by the presen tation of information, and that learning improves when the same information is delivered in different formats (words, pictures, sounds). A message is a display of text and graphics or other signs produced expressly to modify the psychomotor, cognitive, or affective behavior of people. Design refers to the process of analysis and synthesis that starts with an instructional objective and finishes with a blueprint for that objectives realization (Reigeluth, 1983). M essage s presented as an organized set of ideas promote understanding (Witt, 1981). The study of instructional message design allows web designers to make effective choices from a myriad of approaches (Fleming & Levie 1993). Message design principles can be best utilized as guidelines fo r the display of a visual message (Pettersson, 1999). While instructional design principles and practices appear to be critical to successful information processing and acquisition outcomes, the specific focus of instructional message design is linking th eory and practice (Fleming & Levie 1995). It seems logical to use effective

PAGE 22

12 12 message design strategies to impart information to web site visitors in the field of tourism, where technology can be used to promote message retention enhance commu nication, and provide sources of further information Evaluation of web sites according to instructional message design principles is a method of appraising web site effectiveness, but, to date, this method has not been used. Theoretical Framework This s tudy is based on the theory of cognitive load and the role it plays in information acquisition and processing. Cognitive load refers to the demands on working memory (Sweller, 1994; Sweller, Van Merrinboer, & Paas, 1998). According to cognitive load the ory, the limitation of working memory capacity is a key issue in the learning process. To maximize retention content should be presented in such a way that the learners working memory is focused on relevant information and is not taxed by irrelevant or distracting features of material. Information acquisition and processing should be supported. Instructional message design principles provide guidelines for developing effective messages. Use of supportive message design elements, such as effective text a nd graphics, increases attention and motivation, which, in turn, reduces cognitive load. Reduced cognitive load enhances the processing of the intended message into long term memory. Figure 1 2 illustrates these relationships. By providing information th rough supportive message design, CVBs can enhance the effectiveness of their message delivery Effective use of text and graphics can promote more efficient consumer information acquisition and

PAGE 23

13 13 processing. This study examines use of supportive message de sign elements on CVB web sites. Figure 1 2. The relationship supportive message design, cognitive load, and information processing. Purpose of the Study The use of effective message design strategies can reduce cognitive load on working memory and facilitate retention in long term memory (Flem ing & Lev ie 1995) It is clear that information providers in the travel and tourism i ndustry can benefit from using message design strategies that enable their message be retained in the long term memories of their web site visitors. The purpose of this study was to examine destination message designs of informational (CVB) web sit es in travel and tourism through the prism of instructional message design. This study assesse d the use of instructional message design elements that can promote the acquisition and retention of CVB destination messages by web site users through reductio n of cognitive load

PAGE 24

14 14 The goal of this study wa s to assess destination message design on travel and tourism web sites and determine what effective message design practices are used to disseminate tourism related information across world regions More specifically, this study was conducted to answer the following research questions: 1. What is the prevalence of message design elements on CVB web sites? 2. Which message design elements are CVB web sites using effectively to deliver destination messages? 3. How many underlying instructional message design conditions that enable the desired retention outcomes are present on CVB web sites? 4. Do regional or geographic differences exist in the use of message design elements on CVB web sites? 5. What web page elements do web design professionals believe to be most important for destination message delivery? In Chapte r 2, a review of related literature is presented to provide theoretical and empirical support for the message design elements examined in this study and their role in tourism information search. An overview of the methods used in this study to answer thes e research questions is provided in Chapter 3. The results from the study are presented in Chapter 4. Finally, in Chapter 5, the implications of the results for consumer information acquisition and processing theory are presented, along with recommendati ons for future research and for message design strategies for CVB web sites.

PAGE 25

15 CHAPTER 2 REVIEW OF THE LITERATURE The personal computer, initially introduced in 1979, required a number of years to become established as standard business equipment. As c osts decreased, computer ownership became more widespread. By the end of the twentieth century, more personal computers than televisions were sold in the United States each year (Kormier, 1999). The World Wide Web In its relatively short lifespan, the Wo rld Wide Web has grown tremendously. Retail sales generated from Internet commerce were in excess of $74 billion in 2002 (Rosen & Purinton, 2002). The growth of the Internet in sheer number of users is changing so quickly that it is impossible to determin e the population of users at any given point in time (Kormier, 1999). For most companies, the question is no longer whether or not to have a web presence, but rather how to produce the best site for their particular business (Kovacs & Rowell, 2001). Main taining a beautifully designed, current, and informative web site can establish a companys presence as a major player in the field and enhance the identity of the company (Kormier, 1999). Vogt and Pennington Gray (2002) demonstrated the growing impact of new tourism information sources such as the Internet, as well as decreased uses for traditional tourism marketing approaches.

PAGE 26

16 Web Site Design A web site can be defined as a number of content elements, or pages, that are linked together. The essential fea ture of hypertext design is that it provides options for deep, nonlinear searches for information that are under the control of web site visitors. Web designers can influence the outcomes of hypertext searches by imposing a structure on the web s ite, that is, by providing links between content elements (Huizingh, 2000). Web sites often replicate the content and structure that mirrors the needs or concerns of the organization rather than the needs of users. If a web site does not meet the expectat ions and needs of its visitors, then it will not meet the needs of the organization represented on the site. Web site development must be user centered, and the needs of the user should be paramount when maintaining or making changes in web site design (Be van, 1998). As technology progresses, web sites continue to evolve, becoming faster, richer, and more functional, with increasing numbers of visitors using them to do a myriad of things (Motley, 2000). Web site users have more choices than ever, and may b e unwilling to waste time on sites that are slow, confusing, or do not satisfy their needs. As a result of near infinite choice and the ease of browsing, web users exhibit singular impatience and demand instant gratification (Raward, 2001). Long waiting times, or what web users perceive to be long waiting times, are not well tolerated in the online environment. The problems associated with long waiting times are problems for web designers as well, because web sites

PAGE 27

17 whose waiting times are routinely long (greater than two seconds) are visited less frequently than others (Fui Hoon Nah, 2004). Jung and Butler (2000) listed some important characteristics of a successful web site, including appearance, ease of navigation, design, strategic partners, content interactivity, promotion, feedback, repeat visits, value added, currency, and message. Gibson (2001) advises that a web site should be interesting, fun, interactive, and user friendly, using a fair amount of graphics, graphic icons and interactive links to develop feedback and user participation. Many site visitors will expect two way communication (Motley, 2000), such as email, bulletin boards and chat rooms. Incorporating specific features into a web site serves to improve customer service, log demog raphic information, open new markets without respect to geographical constraints, increase revenues, and reduce the number of incoming telephone calls and letters (Kormier, 1999). The complexity of web site design involves mastery of a wide scope of k nowledge: (a) knowledge of the communication process; (b) understanding of hardware, software, graphics, animation, video, and sound; (c) appreciating evolving issues, characteristics, and capabilities of the Internet; and (d) analysis, design, implementat ion, and maintenance of web sites (Kovacs & Rowell, 2001). In addition, designers need to understand the goal of the web site, whether it is to sell a product or to inform or educate a site visitor. Instructional Web Pages The World Wide Web is now the pr imary technological media for delivering instructional material; indeed, a primary goal of many web pages is

PAGE 28

18 educational in nature. When creating educational or instructional web sites, web developers may use general purpose methods, similar to those metho ds used for creating conventional web pages. The problem with this practice is that it does not address requirements that are intrinsic to instructional web sites. That is, they do not incorporate those aspects that are specific to education and the teac hing learning process. Developing an instructional web site necessitates the inclusion of instructional, structural, functional, and aesthetic properties (Montilva, Sandia, & Barrios, 2002). The primary responsibility of an instructional or informational web site is to create a place or an environment where acquisition and retention of information can occur (Hubbard, 1998). W eb pages must be well designed to provide an environment that can enable a successful information acquisition experience. Hype rlinks provide a form of user control over content and enable the web site visitor to interact with the content. H yperlinks allows users to go deeper into subject matter using the web as a convenient electronic library (Morrison & Guenther, 2000). Informat ional web sites can be made more effective through strategies offered from the study of instructional message design. Some researchers in web development have offered prescriptions for web site design that are reflective of tenets in the field of instruct ional message design. McFarland (1995) states that at each step of development the developer should determine whether text, illustrations or icons should be used. Although the presentation of

PAGE 29

19 information can be dominated by text, the option to use pictures or text or some combination must be considered. Web sites often contain text based materials that were created in a traditional print format, but have not been adapted for presentation on the web. Web site visitors rarely read web pages wo rd for word. Instead, they scan to find the information they need. Web designers should make text easy to scan by using meaningful headings, bulleted lists, and highlighted key words. Further, large sections of text should be available for printing or down loading, as most users will not read large amounts of text online (Bevan, 1998). Designers should also e nsure that the intended message is not skewed by the packaging of the information. Many multimedia packages deliver more information than users can p rocess at one time, thus increasing cognitive load and hindering the educational process. To protect against learner frustration, the q uantity of information presented should be limited to digestible bites (McFarland, 1995). Color media is preferable to b lack and white (Simonson, 1984). Color enhances communication when properly used. When color is not properly used it may tend to confuse or even offend. Designers should use a consistent color scheme. Soft nonintrusive background pastels and soft grays are said to provide the best backgrounds, as the human eye may become fatigued after long exposure to highly saturated colors (McFarland, 1995). Bevan (1998) suggests that p atterned or multicolored backgrounds can make text difficult to read There are two major problems inherent in web use for information retention : (a) loss of overview as a consequence of limited s creen area, and (b)

PAGE 30

20 the shortened attention span of the learner due to fatigue from reading online. Loss of overview occ urs when the volume of information that can be presented on screen is limited compared to print media. To mitigate loss of overview, web content should be presented in a generalized format initially, with details to follow. Only meaningful information sho uld be included (Weitl, S, Kammerl, & Freitag, 2002). Nine considerations for designing informational web content were listed by Weston, Gandell, McAlpine, and Finkelstein (1999), including computer literacy, computer access, institutional and infrastr ucture interactivity, navigation, currency and validity of information, the technological considerations of loading speed, and bandwidth Table 2 1 lists various web site elements that have been recommended for web sites to optimize information acquisitio n and retention Table 2 1. Support in the literature for web site content elements Web Site Content Elements External Validity Currency of content Govers et al. (2000), Jung & Butler (2000), Motley (2000), Innovations in Distance Education. (1998), Pan & Fesenmaier (2000), Raward (2001) Date of last update indicated Jung & Butler (2000), Kormier (1999), Motley (2000), Pan & Fesenmaier (2000), Raward (2001) New information indicated Motley (2000), Raward (2001) Purpose/mission statement Gretzel et al. (2000), Jung & Butler (2000), Kovacs & Rowell (2001), Raward (2001) Table of contents Jung & Butler (2000), Raward (2001) Color pictures Gibson (2001), Schuman (1996)

PAGE 31

21 Table 2 1. Continued. Web Site Content Elements External Validity Navigation tools on all pages Govers et al (2000), Gretzel et al. (2000), Jung & Butler (2000), Motley (2000), Poling (1999), Raward (2001), Schuman (1996), Tierney (2000) Site map Motley (2000), Raward (2001) Main message of the page presented before the fold Schuman (1996) Text equivalent for non text elements Raward (2001), Schuman (1996) Information conveyed with color available without color Raward (2001) Minimum of one link per page Fesenmaier (2000), Poling (1999), Pan & Raward (2001), Schuman (1996) Backgrounds Jung & Butler (2 000), Schuman (1996) Navigation bar Poling (1999) Tables Kovacs & Rowell (2001) Animated GIFs Gibson (2001), Poling (1999), Schuman (1996) Text effects Innovations in Distance Education (1998) Clip art Schuman (1996) Graphics Poling (1999) Links to further (internal) resources Tierney (2000), Innovations in Distance Education. (1998), Raward (2001) Reliability of links to external resources Govers et al. (2000), Innovations in Distance Education. (1998), Raward (2001) Frequently Asked Questions (F AQ) included Jung & Butler (2000), Govers et al. (2000), Kormier (1999), Innovations in Distance Education (1998), Raward (2001) Internal search tool Govers et al. (2000), Gretzel et al. (2000), Pan & Fesenmaier (2000) Affiliate program Jung & Butler (20 00), Gretzel et al. (2000), Poling (1999) Security Jung & Butler (2000), Poling (1999), Pan & Fesenmaier (2000)

PAGE 32

22 Table 2 1. Continued. Web Site Content Elements External Validity Can user comments be recorded? Raward (2001) Is it possible to get fe edback? Gibson (2001), Innovations in Distance Education (1998), Raward (2001) Is it possible to ask questions? Motley (2000), Innovations in Distance Education (1998), Raward (2001) Contact information available Motley (2000), Raward (200 1) Help available Innovations in Distance Education (1998), Raward (2001) E mail Kormier (1999), Gibson (2001), Motley (2000), Gretzel et al. (2000), Innovations in Distance Education (1998), Pan & Fesenmaier (2000), Raward (2001) Bulletin board Gretzel et al. (2000), Innovations in Distance Education (1998), Pan & Fesenmaier (2000), Raward (2001) Chat room Gretzel et al. (2000), MacDonald & Caverly (2001), MacDonald & Caverly (2001), Innovations in Distance Education (1998), Pan & Fesenmaier (20 00) Guest book / visitor comments Pan & Fesenmaier (2000), MacDonald & Caverly (2001) User forum (threaded) MacDonald & Caverly (2001), Innovations in Distance Education (1998) Forms Pan & Fesenmaier (2000) Fax Innovations in Distance Education (1998) Java Poling (1999), Pan & Fesenmaier (2000) Listserv Innovations in Distance Education (1998) Flash presentations Pan & Fesenmaier (2000) Headings stand out on the page Raward (2001) Reading level appropriate Innovations in Distance Education (1998), Raward (2001) Page scan ability Raward (2001)

PAGE 33

23 Table 2 1. Continued. Web Site Content Elements External Validity Text simple, concise, clear Govers et al. (2000), Innovations in Distance Education (1998), Raward (2001) Visual appeal Jung & Butler (2 000), Schuman (1996), Gretzel et al. (2000) Format consistency Kovacs & Rowell (2001), Schuman (1996), Raward (2001) Information presented in readable blocks Innovations in Distance Education (1998) Headings brief and informative Raward (2001) Tech nology in Tourism Tourism destinations are offering web sites that provide travelers with online travel brochures to assist in planning business or pleasure trips. Govers, Jansen Verbeke, and Go (2000) assert that the tourism industry is too focused on ac tual booking and sales, when actual booking is only one phase of the decision to purchase process. Use of the web in information gathering and planning may be more advanced than for actual booking and payment (Govers, Jansen Verbeke, & Go, 2000). The ava ilability of current, up to the minute information is a primary advantage of travel planning through the Internet (Schley, 1997), as is the use of multimedia. Multimedia is an umbrella term that refers to using more than one form of media in a single medi a product (e.g., print, sound, animation). Multimedia encompasses a wide range of technologies, and is considered the technological wave of the future (Sigala, Airey, Jones, & Lockwood, 2000). Interactive media is especially suited to the information int ensive travel and tourism industry (Gretzel,

PAGE 34

24 Yuan, & Fesenmaier, 2000), where content delivery must relate to the knowledge base of the user. Building a bridge between presented material and the visitors existing knowledge base is a key factor in success fully using multimedia. As a rule, a person will learn more if many links are established by the instructional content to prior experience and knowledge. The establishment of these links is known as information mapping (McFarland, 1995). Information Searc h The World Wide Web is an important tool for information search and retrieval (Fui Hoon Nah 2004). When making a purchase decision, the tourist may consult many sources of information. This process is known as information search. Consumers search for i nformation more so when product costs are high (Dana, 1994). The types of information sources available are major factors in a persons decision to purchase a specific tourism product or service. Tourism products and services have inherent uncertainty asso ciated with them. That is, tourism products and services are usually expensive and cannot be previewed or sampled prior to purchase. Therefore, the tourist who travels to a destination for the first time commonly conducts information search using several information sources, including both marketer dominated promotional materials and objective consumer information services (Eby, Molnar, & Cai, 1999). Most purchase decision making models feature pre purchase information search as a key factor (Bieger & Lae sser, 2004). A good understanding of information search behavior is key to planning and decision making for businesses (Moorthy, Ratchford, & Talukdar, 1997 ). There is a cost associated with each consumer alternative, and

PAGE 35

25 that cost may be either time spent or cognitive effort (Meyer, 1982). Cognitive costs, such as the evaluation of information and consideration of alternatives are also integral to information search, although the consumers perceptions of these costs are subjective and may vary depending on the searchers mood, involvement, or education. For this reason, the total cost of information search to consumers is incalculable. Information search activity inevitably consumes the searchers time and energy, in addition to the financial costs asso ciated with online access (Hodkinson & Kiel, 2003). As a result, we should expect information search to be utilized only by those people for whom the value of information search exceeds the cost (Balakrishnan, 1991). Vogt and Fesenmaier (1998) considered information search behaviors in the travel and tourism industry, and expanded the role of tourism information search from a primarily marketing purview into a broader communication context. Consumers learn from their own experiences, and through visual, v erbal, and sensory stimuli that educate them about products, including tourism destinations. This information is then stored in memory and used for functional reasons such as planning or taking trips ( see Figure 1 1). Major factors identified in the tourism literature which influence information search include price, the composition of the vacation group, the presence of family and friends at the destination, prio r visits to the destination, and the degree of novelty associated with the destination (Snepenger, Meged, Snelling & Worral, 1990). The distinguishing feature of consumer information search behavior is that of consumers prior knowledge of the market and any implications of this

PAGE 36

26 knowledge for the search process. Advertising can influence a consumers brand perception and method of information search ( Baloglou, & McCleary 1999; Gallarza, Saura, & Garcia, 2002 ; Moorthy, Ratchford, & Talukdar, 1997). Wilkie and Dickinson (1985) state that an important research goal is a fuller understanding of the benefits consumers achieve by information search behaviors, along with more in depth details of the search process itself. In a study centered on the multiple info rmation needs of tourism information seekers, Vogt and Fesenmaier (1998) found that information is primarily collected and used for functional reasons, such as planning or taking trips. Other, secondary needs surfaced in the forms of hedonic and aesthetic needs that play a critical role in the information search process. Hedonic needs allude to the view of consumers as pleasure seekers and the need for information search to produce some pleasurable effect on the searcher, such as enjoyment, amusement or s ensory stimulation. Aesthetic needs can be satisfied from both verbal and visual forms that stimulate visual thinking and imagery, along with fantasizing about places that seem real and obtainable. They concluded that, to guide potential tour ists toward a purchase decision, tourist information should appeal to a variety of search needs to first capture the attention of potential travelers. The model of consumer purchase behavior adapted by Vogt and Fesenmaier (1998) depict the role of active a nd passive external sources in information acquisition and processing. A key external source for travel and tourism information acquisition is a destination web site. Current technologies have enabled web sites to act as both passive and active sources o f information.

PAGE 37

27 Web site visitor interest can be engaged by utilization of content and support elements that act as passive information presentation strategies. Information search behavior is an important concept in the study of consumer behavior, and th e web provides a major search resource for consumers (Hodkinson & Kiel, 2003). In a study that profiled people who used the Internet to gather travel information or purchase travel arrangements, Weber and Roehl (1999) found that travel information searche rs were 26 to 55 years of age, had higher incomes, were in professional or computer related occupations, and tended to have more online experience. In terms of time spent, the Internet was gaining rapidly on more traditional types of media like television and newspapers Understanding the process by which consumers acquire information is critical for management and marketing decisions (Bieger & Laesser, 2004). The motivation for modeling how consumers gather information is to better understand the cognitive processes governing consumer behavior in the marketplace (Meyer, 1982). The goal of information search is usually to obtain a specific bit of targeted information and recall of the full content of information contained in the sear ched document is generally not the focus of the reader. As a result, the cognitive processes needed for limited information retrieval are utilized to a greater extent than those needed for prose comprehension (Guthrie, 1988).

PAGE 38

28 Incorporating Instruction al Message Design into the Consumer Information Search and Processing Model Digital technology and interactive multimedia have had a significant impact in the field of education, where web based learning environments utilize the power of the Internet to c reate a context where learning is both supported and enabled. The Internet offers many unique features for learners and can provide a wealth of information that is not readily available elsewhere. Furthermore, the information available on the web is usua lly current, presented in meaningful contexts, and linked to other available information that explores topics more widely. In addition, the Internet affords opportunities to be more interactive and collaborative, creating a true global community where tec hnological tools such as email, listservs, newsgroups, and videoconferencing enable exchange of knowledge, ideas, and perspectives. Development of the Internet and high speed communication systems has led to new opportunities for delivering information (Te h, 1999). In the field of education, digital technology can provide a means to utilize teaching methods that have been demonstrated to be effective, providing ways to engage students in active learning and offer easier access to vast amounts of informatio n. Evidence has revealed that when people are actively involved in a self driven learning project, they learn more and remember it longer than when they are passively sitting and listening (Newman & Scurry, 2001). The challenge for web

PAGE 39

29 designers is not o nly to generate traffic to their web sites, but also to keep visitors involved using proven design practices (Jung & Butler, 2000). In the digital environment, technology has already begun to alter the way people learn in every setting from skill training sessions to elementary, high school and college courses, as well as in traditional classrooms (Newman & Scurry, 2001). Learning behavior theorists believe that adult learning attitudes and behaviors are built upon childhood learning experiences. Consumer s who are obtaining information from the web do so by exhibiting previously held learning behaviors in a new and technically mediated environment. The process of information search, evaluation and decision making thus involves learning and learning styles (Hodkinson & Kiel, 2003). Therefore, benefits from learning strategies that have been proven effective in online classrooms may transfer to instructional web site design. In either venue, technology will be used to enhance learning, encourage communicat ion, and provide sources of further information (Newman & Scurry, 2001). Effective teaching and e learning strategies have been widely reported. A summation of some effective distance learning strategies was published in Innovations in Distance Education (1998). This included the following suggested strategies: 1. A comprehensive 24 hour a day, 7 day a week system of support in place. 2. Regular feedback mechanisms maintained. 3. Frequent and meaningful interactions among learners, between learners and instruction al materials, and between learners and the instructor. 4. E communications technologies should be used as a tool for creating and maintaining learning communities.

PAGE 40

30 5. Programs should employ creative solutions to achieve the traditional outcomes desired. 6. Social interactions between and among e learners enrich the learning community and should be supported in the design and delivery of the program. 7. The program should incorporate a technology base that is appropriate for the widest range of e learners within the t arget audience. 8. The selection of instructional media and tools should reflect a thorough understanding of the added value of the technology. 9. The design should take into account the diversity of potential users. In the virtual classroom, the emphasis is on the learner or on how the learner is going to assimilate the material. People have various learning styles, and that fact must be taken into account. For online learning to be successful different learning styles must be accommodated (Weitl et al. 2002). An interactive environment should also be fostered (Gibson, 2001). Important factors for successful marketing on the Internet include attracting users, engaging interest and participation, retaining users, learning about user p references, and providing feedback (Gretzel, Yuan, & Fesenmaier, 2000). Metros (2001) found that online learner engagement ranges from passive interest to dynamic interaction to flow. The term flow refers to states of deep, intense involvement in challeng ing activities that do not overwhelm the skills of the learner (Csikszentmihalyi, 1975). Web site browsing, for example, would demonstrate passive interest by the site visitor. The tools of interactivity could then be used to stimulate dynamic interactio n where a flow state may eventually be achieved. Use of interactive web site features such as threaded discussions, chat, and live web cams can act as active external information sources, and at the same time provide the dynamic interaction necessary to e ngage the site

PAGE 41

31 visitor. The information content should be educational and should be written to inform people about attractions, destinations, and services (Eby, Molnar, & Cai, 1999). Cho and Fesenmaier (2000) use the term telepresence to describe much th e same process. Telepresence is akin to Csikszentmihalyis (1975) concept of flow because both states involve a high degree of involvement in a task or activity to the exclusion of outside stimuli or the awa reness of self or time. Internet Web Site Evalu ation Evaluation of web site effectiveness is necessary, in part, due to the substantial costs associated with establishment and maintenance of web sites (Tierney, 2000). In general, the web lacks standardized measures for evaluation, as the complexity of the web makes standardization a challenge (Gretzel, Yuan, & Fesenmaier, 2000). Tabulating direct sales from a site is one method of evaluation, but many DMO web sites do not sell directly to the consumer and cannot be measured in this manner. Further, it is well established that visitors to a site are often there primarily to gather information. It is therefore critical that the information that a consumer seeks be presented in a clear and memorable fashion if later sales are to be realized (Tierney, 20 00). A common method of measuring usage of web sites is number of hits, followed by number of user sessions, and online feedback from site users. Tracking hits as a measure of web site effectiveness can be misleading,

PAGE 42

32 however, as the amount of traffic t o a site may bear little relation to the effectiveness of the site (Tierney, 2000). The nature of web sites allows designers to easily and effectively log site usage, number of hits, and patterns of site navigation. This information, however, reveals l ittle about visitor characteristics, motives, preferences, satisfaction, or likely actions that the visitor will take as a result of visiting the web site. Incorporating interactivity into the web site, however, can enable the visitor to impart a great de al of personal information back to the host (Tierney, 2000). Jung and Butler (2000) listed some common measurement methods that determine the success of a web site, including number of hits, time spent, booking rate, interactivity, repeat visits, feedback currency and gateway partnerships (commercial banners). A more reliable evaluation statistic is the user session defined as a session of activity (all hits) for a particular user at a particular site. User session information can be generated by soft ware packa ges and provide data on visitor s domain names, region or country of request, organization type, advertisements viewed, and time spent in various areas of the site. However, no information on visitor demographics, motivations or likely actions t aken from viewing the site can be known from user session type data (Tierney, 2000). An evaluation method growing in popularity is the online survey, a interactive method that can provide detailed data on the web site visitor. A serious concern about I nternet surveys, however is nonresponse bias. A survey

PAGE 43

33 can only measure data from those who choose to participate, and Internet surveys can have abysmal response rates (Tierney, 2000). Another site effectiveness or tracking strategy employed by commer cial web sites is the use of cookies, which are data files used by web servers to identify web users and track user's browsing habits (CIAC, 1998). Unfortunately, cookies cannot identify individual users, and many web surfers simply turn off the accept cookies option in their web browsers to avoid tracking (Tierney, 2000). According to Motley (2000), o ne of the most crucial characteristics of a good web site is freshness the driver of web site stickiness To retain customer lo yalty, Internet based companies and web sites seek stickiness the ability of a site to attract repeat customers. People expect to interact with a web site to experience new things or learn current information. Offering some type of interaction e nsures th at there is a reason to return to a web site Similar descriptions of types of web site designs were offered by business (Poling, 1999) and educational (MacDonald & Caverly, 2001) researchers (Table 2 2 ). For education, the primary concern is to select the most appropriate technology for the teaching and learning situation. Pedagogy comes first; technology is only utilized to assist the teaching (MacDonald & Caverly, 2001). Given the mission of the CVB as both a marketing and educational entity, elements from both business and education should be considered. Instructional Message Design Tufte (1990) noted that the principles of information design are universal and are nonaligned with any particular language or culture. The specific focus o f

PAGE 44

34 Table 2 2 Web site evolution by generation Poling (1999) MacDonald & Caverly (2001) Generation 1 Web Site Simple online presence, an electronic business card Provides online information with links to further sites Generation 2 Web Site Polished, professional look, quick and easy navigation, and clear goals for results Incorporates G1 information, but adds interactivity through email and web based forums Generation 3 Web Site Cutting edge marketing techniques such as affiliate programs, interacti vity, and using animation and Java programming to enhance functionality and appeal Includes features of G1 and G2, but introduces synchronous interactivity through audio/video conferencing and chat rooms instructional message design is linking learning theory and educational practice (Fleming & Levie 1993). The basic tenets for linking educational research and practice are instructional message design principles that establish links between instructional conditions and learning outcomes. There is a difference, however, between descriptive and prescriptive principles. Both involve instructional conditions and outcomes. Prescriptive principles, however, expressly describe learning processes, whereas descriptive principles describe methods of instr uction. Prescriptive principles are clearly a greater value to instructional developers, because they link instructional methods with learning outcomes. Positive outcomes for message design, then, are expressly enabled by principles that illuminate those instructional methods that are most likely to result in the instructional designers intended outcomes (Fleming & Levie 1993). The essential instructional message design question is

PAGE 45

35 this: what conditions should be created that are most likely to resu lt in the desired outcome (Fleming & Levie 1993)? Reigeluth (1983) defines instructional design as the process of deciding what instructional methods are best for enabling intended changes in knowledge or skills. A message is a display of text and graphics or other signs produced expressly to modify the psychomotor, cognitive, or affective behavior of people. Design refers to the process of analysis and synthesis that starts with an instructional objective and finishes with a blueprint for that ob jectives realization Message design principles can be best utilized as guidelines for the display of a visual message (Pettersson, 1999). Winne (1995) suggested that learning is affected by the presentation of information, and that lea rning improves when the same information is delivered in different formats (words, pictures, sounds). The study of instructional message design allows designers to make useful choices from a myriad of instructional methods and materials to promote informat ion acquisition and retention (Fleming & Levie 1993). Complicated language, texts, pictures, or graphic design will hinder the learner in understanding the intended message. Informational text and graphics should be designed so that they are easy to read. The goal of information design is clarity of communication. For learning to occur, however, the learner must be interested and curious about the subject matter (Pettersson, 1999). Learning increases when learners attitudes are more positive (Simo nson, 1984). Message designers must ensure that the intended message is not skewed by the packaging of the information. Many multimedia packages deliver more

PAGE 46

36 information than users can absorb at one time, thus hindering the educational process. To minim ize learner frustration, the quantity of information presented should be limited to digestible bites (McFarland, 1995). Impositions on working memory from nonlearning events can also be reduced to the effective use of different media. Still another strate gy to reduce the imposition of short term memory caused by complex message comprehension is to make instructional messages simple and easy to understand (Farquhar & Surry, 1995). For delivery of content in online learning, the structure and presentation of content should be similar to the structure and presentation of hypermedia on the Internet According to Weitl, S, Kammerl, and Freitag (2002), t here are two major problems inherent in using the web for learning: (a) loss of overview as a consequence of limited screen area, and (b) the shortened attention span of the learner to fatigue from reading online. Loss of overview occurs when the volume of information that can be presented on screen is limited compared t o print media. To mitigate loss of overview, web content should be presented in a generalized format initially, with details to follow. Only meaningful information should be included Understanding of instructional message design is incomplete without understanding the role of motivational appeal. The ultimate success of an instructional message will depend upon both the learner's willingness and ability to receive and understand the message (Fleming & Levi e 1993). In the case of CVB web sites, designers must consider how the destination message can be delivered in such a way that will motivate the user to seek additional information.

PAGE 47

37 Goals of Effective Message Design Effective, supportive message design s hould enhance information acquisition and retention by reducing cognitive load, or what Miller (1956) would describe as the restraint on the amount of information we are able to process and remember. Design elements can enhance or inhibit motivation and a ttention. Sound message design always serves to maximize motivation and attention, while eliminating elements that detract from these goals. Motivation. The primary goal of instructional message design is to utilize principles of motivation that result i n increased motivation to learn. Gagn (1985) lists motivation as the first event of instruction, but instructional designers should consider motivational appeal in every part of the message. Motivation can be defined as that characteristic which determine s the magnitude and direction of behavior. In this definition, magnitude is the degree of effort, while direction can be defined as optimum orientation toward goal (Keller & Burkman, 1993). Motivation in learning is dependent on the learner's personality and learner's perception of value and the difficulty of learning objective. Instructional message designers confront this reality by accepting, in some part, responsibility for motivation in learning. While it is true that an instructor does not totally control a learners motivations, poor instruction can demotivate just as great instruction can inspire learners (Keller & Burkman, 1993). Attitudes are also related to learning and can be changed by technologically mediated instruction (Simonson, 1984).

PAGE 48

38 F or the learner, motivation is a primary determinant in the outcome of learning events. For this reason, motivation should be considered throughout the instructional message design process. The specific focus of instructional message design is linking lear ning theory and educational practice. The study of instructional message design allows educators to make useful choices from a myriad of instructional methods and materials (Fleming & Levie 1993). Learners are more likely to feel confident and motivat ed about instructional materials that convey a favorable first impression than those that do not. Most motivational text and graphic principles relate to gaining and maintaining attention, relating content to learner interests, goals, or past experience, a nd building and maintaining learner confidence in the ability to use the content (Keller & Burkman, 1993). We can assist the learner in understanding the message by presenting it as an organized set of ideas (Witt, 1981). The organization and delivery of the message affects the perceived credibility of the message (Simonson, 1984). According to Weitl et al. (2002), to mitigate loss of overview, web content should be presented in a generalized format initially, with details to follow. Only meaningful information should be included. To mitigate the probable short attention spans of the learners, web content should be presented in small volumes Messages that tell a story have a greater persuasive impact than mess ages that simply provide information. When information is structured and presented logically and intelligently, it is more likely to be perceived as persuasive. One effective technique that facilitates an increase in perceived

PAGE 49

39 credibility is to use actors demographically similar to the target audience (Simonson 1984). Learner attitudes are changed when mediated messages are germane and technically well done, rather than irrelevant and poorly produced. Mediated instruction that is perceived as realistic rel evant and technologically stimulating is more persuasive to the learner (Simonson, 1984). In many studies concerning use of media and instruction, achievement, and not motivation, has been the dependent variable. Other media research has focused on the im pact of short messages with limited content, such as TV commercials and print advertisements, and therefore does not equate easily to instructional design (Keller & Burkman, 1993). Motivation plays a critical role in the CVB mission of destination message delivery; a CVBs web site must motivate the visitor to seek additional information. Attention Industrialized countries are also mass media societies. Living in a mass media society means exposure to a constant stream of information via media that pe rmeates home, school, and work environments. It is nearly impossible to avoid this constant stream of information. But, in this sea of information bytes, it may also be very difficult to obtain the specific information that we really seek. Because audio text and visuals constantly compete for our attention, it is possible to miss the information that we are really interested in. Our attention is directed not only to those topics that we are interested in, but also to sounds, to objects that move, to ob jects that are large, bold, new, or have the right color, shade or contrast that enable them to stand out from the environment (Pettersson, 1999).

PAGE 50

40 Urbina (1994) defined attention as readiness on the part of the receiver to perceive stimuli in the environme nt. Attention determines which environmental events we become conscious of. Gagne (1977) averred that gaining the attention of the learner was the first step in instruction. In similar fashion, the first problem for an information designer is to gain the attention of individuals. Informational material must capture the attention of the intended audience. In addition, the message must be understandable to the intended audience ; that is, the content of the message must in some sense agree with the learner 's attitudes and beliefs and motivate the learner to understand the message Attention can be controlled automatically or by specific instruction; an information designer can select from different guidelines and design principles in orde r to affect the reader's attention (Pettersson, 1999). Attention is the mechanism that determines which stimuli will be engaged and which stimuli will be ignored (Carlson, 1993). Mayer (1993) stated that for learning to occur, the learner must first pay at tention to the relevant information. To accomplish this goal, instructional designers must effectively draw the learner's attention to the relevant information. Attention is never objective; it is always a selective process (Pettersson, 1999). Pashler ( 1995) noted that full attention enables the fast and efficient completion of tasks. Much of attention is controlled by will, a conscious internal process. External stimuli can also affect control of attention. External stimuli that are new or intense ca n sometimes control our attention for short periods of time. If the same external stimulus is repeated or becomes continuous, its command of attention will diminish. The

PAGE 51

41 use of effective message design principles can enhance the CVB web sites capacity t o get and maintain the users attention. Visuals and graphic images are perceived more rapidly than texts (Fleming & Levie 1978). An often repeated opinion on the function of visual images concerns getting attention and maintaining attention (Petterss on, 1998). Among the visual devices with attention getting capabilities are split screens, shading and contrast, text and graphics, and zoom lens type movements that illuminate or emphasize specific details. Although pictures and graphics may not carry t he main weight of a message, they are useful in attracting the direct attention of the learner (Witt, 1981). The primary function of still pictures is to attract and hold attention, to facilitate, persuade, illustrate, clarify, motivate or reinforce (Pett ersson, 1998). Fleming and Levie (1978) stated that an appreciation of the movements of the human eye is important to message designers understanding of getting and maintaining attention. We scan the things we look at (Pettersson, 1999). Bevan (1998) argues that web site visitors rarely read web pages word for word. Instead, they scan to find the information they need. Message designers can make texts scanable by using meaningful headings, bulleted lists, and highlighted key words. Bevan also cautions that unnecessary whitespace may make scanning of the text more difficult, resulting in an imposition on gaining attention. Such impositions result in non optimal learning environments that can lead to shortened attention spans among learners (Wei tl et al. 2002).

PAGE 52

42 It is important for designers of text based content to design text that creates a favorable first impression. A favorable first impression enables potential learner's to scan the content and to feel confident t hat they can easily read and understand messages. A favorable first impression or image may help to gain and keep learner attention and build learner confidence that they can successfully navigate and use the content (Keller & Burkman, 1993). It is advisa ble, therefore to make the initial perception of printed content easy to read and use in order to build confidence and maintain attention. According to Keller and Burkman (1993), well organized and explicit text also maintains learner attention and bui lds confidence Additionally, using readable language, a familiar font and font size, and appropriate color graphics helps gain and maintain learner attention To further mitigate probable short attention spans of learners on the web, content should be presented in small volumes (Weitl et al. 2002). Gaining and keeping the attention of the learner is a useful tool for reducing the imposition of nonlearning events on working memo ry. Gagne, Briggs, and Wagner (1998) report that changing stimuli and appealing to learner interests can gain and maintain attention. Message Design Elements Enhance Attention and Motivation The most critical elements of supportive message design are th e text and graphics used to convey the message. Numerous factors contribute to the effectiveness of text and graphics. It is possible to use text and graphics as motivational tools and as methods to get and maintain the learners attention.

PAGE 53

43 The appropria te use of textual features such as font, graphics, and page design can increase attention and influence motivation to learn, even though these methods may be secondary to teaching methods and content (Keller & Burkman, 1993). Text Text is the most commo n medium of instruction. Text comprehension is a constructive process in which the learner integrates the content with his or her own experience (Pettersson, 1999). When designing text, it is useful to consider your target learner's reading skills and int erest in reading. Most principles that are successful with below average learner's also can be applied to above average learner's, although the converse is not true (Keller & Burkman, 1993). Although readability typically refers to the ease with which th e language in text is read and understood (Chall, Bissex, Conrad, & Harris Sharples, 1996), many other aspects of text contribute to its readability. The phenomenal growth of the World Wide Web, coupled with our increased use of word processing programs fo r creating documents, has increased the volume of text that we read from computer screens (Dyson, 2004). When reading a computer screen, many learners scan textual materials prior to reading them. Dyson suggests that Web pages, in particular, are frequen tly scanned rather than read in great detail The pre reading scan is generally focused on identifying the topic of the material and assessing its level of interest. For many typical learners however, this pre reading scan serves a secondar y function to assess the level of difficulty in reading and using the content. The outcome of that assessment can determine whether learner's approach the task

PAGE 54

44 of reading optimistically or pessimistically, and may cause some learner's to abort reading at that point (Keller & Burkman, 1993). The formatting of text based content can determine whether it is perceived as easy or complex Relatively short blocks of text and concise messages tend to suggest fairly easy reading to some l earners, and that initial easy to read perception builds confidence and maintains attention (Keller & Burkman, 1993). Large sections of text should be available for printing or downloading, as most users will not read large amounts of text online (Bevan, 1998). Readability of text is, in part, dependent upon a balance among font, font size, leading (i.e., distance between lines of type), and column width. For this reason, most graphic designers follow traditional typographic rules when constructing text. As a result, readers have become accustomed to similar standards in typesetting, a familiarity that has become a reader expectation (Keller & Burkman, 1993). Thousands of fonts have been developed and most instructional designers agree that some fonts mak e text more readable than others (White, 1987). Common fonts are easier to read than uncommon fonts (Benson, 1985; Tinker, 1963), and, for print, serif fonts are easier to read than san serif fonts (Tinker ). Changes in fonts or style can indicate c hanges in importance or purpose. Font selection and typographical techniques can focus the readers attention on main ideas, key passages or important concepts. Multiple type styles can be visually confusing (Pettersson, 1999).

PAGE 55

45 To make text easier to re ad, maintain user attention and increase confidence, web designers should choose a font that is both widely used (familiar) and legible. Additionally, the selection of a font that is widely used increases the chances that the font will be installed and co mmonly available on most computers (Keller & Burkman, 1993). Designers can create the perception of easy reading by dividing lengthy blocks of text into visually distinct segments (Keller & Burkman ; Smith & Orr, 1985; White 1987). Short sections of text are preferable (Bevan, 1998). As a rule, wider columns of text should have larger font size and increased leading (Keller & Burkman ). Running text should be in lowercase letters, as text that is printed in all caps reduces reading speed (Henny 1981; Poulton & Brown, 1968). Lines of text should be composed of 8 to 10 words in 12 point font to make text easier to read, maintain reader attention and increase confidence. Lines of text should also be presented in a fairly open, rather than constr ained display (Keller & Burkman, 1993). Pettersson (1991) notes that black print on white paper is over 10% more efficient in learning than white text on black paper. The most legible combination of text color and background is black text on a light yell ow background. It is widely accepted that the perception of text based content is enhanced by good structure and organization. When presenting text based content, the terms structuring and organizing refer to the logical sequencing of topics and subtopi cs. Text organizers such as headings and the use of whitespace are devices that content providers use to make the structure of text more explicit

PAGE 56

46 (Keller & Burkman, 1993). The terms structuring and organizing also refer to relationships among content elem ents, such as illustrations and blocks of text, that are readily apparent, so that content elements that are not related are not mistakenly perceived to be related (Keller & Burkman, 1993). Well organized and explicit text maintains the readers attention and builds confidence (Keller & Burkman, 1993). It is evident that explicit text organizers also make a text appear to be easier to comprehend, and this can be useful as a motivating tool in and of itself. Graybinger (1985) found that the more structured and organized text based content appears, the more students seem to prefer it. It is also helpful to illustrate the meaning of text with supportive pictures or other graphics (Witt, 1981). External text structure refers to the devices that focus the rea ders attention on specific parts of the text (Jonassen & Kirschener, 1982). For example, headings should attract attention, be relevant and identify subject matter. Headings indicate the relative importance of items within the text. Typographic queuing r efers to the use of bold, italic or underlining to draw attention to key items in the text (Pettersson, 1999). According to Cisotto and Boscolo (1995), paragraph headings improve learning, but underlining of key information does not. People who write newsp apers, magazines and other mass distributed media generally use a writing style that incorporates the active voice and short, familiar, image provoking words that are targeted to typical readers (Keller & Burkman, 1993). Message designers should keep text messages simple and

PAGE 57

47 closely tied to the supportive graphics (Witt, 1981), and use a readable writing style to maintain reader attention and increase confidence (Keller & Burkman, 1993). Graphics The United States could be described as a visually illitera te society. Visuals are used for marketing splashes and sensationalism and for entertainment. Recall of pictures is better than recall of text (Gagne & Rower, 1969) and visuals and graphic images are also perceived more rapidly than text (Fleming & L evie 1978). Although visuals are generally easily interpreted, they can seldom convey the same detailed informat ion of text Where visuals are vital, visuals should adhere to their role to support and augment textual content (McFarland, 1995). Lester (1995) suggested that the reason visuals are a powerful form of communication is that they elicit both intellectual and visceral responses. In traditional paper publication of text materials, visuals were sometimes sparingly used due to the high co st of producing color pictures. The cost of reproducing color pictures, however, is not a factor with materials that are distributed electronically over the Internet. This advance in technology played key role in allowing visual information to be distrib uted worldwide in a cost effective manner. To achieve the full educational benefits made possible by using pictorial information in teaching and learning, it is necessary for graphic presentations to be done with an understanding of how people learn from graphics (Goldfarb & Kondratova, 2003).

PAGE 58

48 Text must support graphics and vice versa. When both text and graphics convey exactly the same information, users may become bored or distracted. When both text and visuals are employed, designers should ensure that the intended messages are not redundant. Text and graphics should be complementary, offering slightly different views of related information to promote learning (McFarland, 1995). Further, the spatial context of sound instructional design ensures that le arning is enhanced when text is placed nearer, rather than farther, to corresponding pictures and graphics (Mayer, 1993). Illustrations should be used to augment and complement textual content. A lone picture may be inappropriate for presenting detailed i nformation (McFarland, 1995). Text containing relevant illustrations is more conducive to learning than text alone. Accompanying graphics are particularly adept at illuminating spatial relationships described in the text (Goldfarb & Kondratova, 2003). Fl eming and Levie (1978) noted that in graphics containing multiple figures of uniform color or shade, a figure of a different color or shade will attract attention. Even if color does not add important information to a graphic image, it may still facil itate information acquisition and processing by causing the user to pay more attention or have more interest in the image (Pettersson, 1998). Messages must be readable at an appropriate distance, or when seen for short periods of time. For example, a driv er in a rapidly moving vehicle may only have a short time to read a sign. Therefore, the message on the sign must be processed and understood quickly. The meaning of the sign should be

PAGE 59

49 immediately obvious. Where appropriate, color combined with shade an d location can be used to gain attention. Color may also be used to imply meaning. Red, for example, is commonly used in warnings (Pettersson, 1998). Hartley (1987) observed that colorblindness is more prevalent among men than among women, with as many as 8 10% of the male population affected, versus only 1% of the female population. Red green colorblindness is the most common. Unfortunately red and green are also used very often as key colors in symbols and graphics (Pettersson ). Regular and unfam iliar shapes tend to garner more attention than basic or familiar shapes. Most people can readily perceive basic shapes. Shape constancy is our shared tendency to perceive shapes as the same despite changes in distance, angle, and lighting. This constan cy and perception is one reason that basic shapes are common in the design of symbols and icons (Pettersson, 1998). Evans, Watson, and Willows (1987) pointed out that graphics can change from being engaging motivators to engaging distracters. Pictures ma y be ignored when an overload of pictures is provided in support of a single message. Pettersson explains that w hen too many graphics are used, readers tend to ignore some of them, producing an effect that is the opposite of attention The type of picture that is displayed is important in comprehension (Pettersson, 1999). Tufte (1983) stated that graphical competence requires three different skills: substantive, statistical, and artistic. In spite of this, most graphic work is under t he sole control of artistic personnel (Pettersson 1998 ).

PAGE 60

50 Designers should always ask whether or not the illustration or visual matches the text. Pettersson (19 98 ) found clear differences between the intended and perceived messages. One way to mi nimize this difference is to caption pictures to support their intended interpretations. Graphics and illustrations should be carefully chosen to augment the understanding of the text. If illustrations or graphics are unrelated to the accompanying text m essage, readers may become confused (McFarland, 1995). Although pictures and graphics may not carry the main weight of the message, they may attract the primary attention of user Designers should keep text messages simple and closely tied to supportive graphics (Witt, 1981). Cognitive Load Theory Cognitive load theory as described by Sweller and colleagues (Sweller, 1994; Sweller, Van Merrinboer, & Paas, 1998) reiterates the multiple components description of working memory. In addition, a distinction is made between intrinsic and extrinsic cognitive load, where intrinsic cognitive load refers to the innate difficulty of processing the information itself and extrinsic cognitive load refers to the format or methods used in presentation of the information Intrinsic cognitive load, then, is dependent on prior knowledge and experience, among other factors, whereas extrinsic cognitive load can be affected by instructional message design. The value of instructional message design is its ability to decrease extraneous cognitive load and direct working memory resources to the learning process itself by limiting the effect of the presentation features of instructional material (Tabbers, 1999).

PAGE 61

51 According to cognitive load theory, the limitation of working memory capacity is a key issue in the learning process. For instructional designers, content must be presented in such a way that the learners working memory is focused on relevant information and is not taxed by irrelevant or distracting features of instructi onal material. Where both verbal and visuals are presented, the learner must mentally integrate text and pictures simultaneously, resulting in a high cognitive load (Tabbers, 1999). Physical integration of visual and verbal content can decrease cognitive load and increase learning (Chandler & Sweller, 1991). A learning event occurs when a learner is consciously interacting with subject matter. Learning events include mentally organizing, categorizing, visualizing, and repeating information to facilitate its encoding into long term memory (Farquhar & Surry, 1995). The process of mentally organizing, categorizing, visualizing, and repeating information is simplified and aided by instructional message design. Nonlearning events are those interactions that are not beneficial to the learning process. Any distraction from the instructional message that causes the learners attention to stray can be said to be a nonlearning event. Nonlearning events in an education model can be said to be similar to noise in the communication model. Nonlearning events inhibit the processing of the instructional message into long term memory by occupying the limited capacity of short term memory (Farquhar & Surry, 1995). Sound instructional message design minimizes the occurre nce of nonlearning events.

PAGE 62

52 Message comprehension events occur when learners use their senses to receive information from an audio, video, or text based instructional message. We can define a typical learner as one in the mid range of motivation, neither over or under motivated (Keller & Burkman, 1993). Message comprehension events compete with nonlearning events as the focus of short term memory. As the short term memory requirements for message comprehension increase, processing into long term memory d ecreases. Effort required to comprehend a complex message, or attention distracted by nonlearning events, will impede processing of the instructional message. That is, the combination of nonlearning events and message comprehension events impede learning events (Farquhar & Surry, 1995). Central to many models of cognitive processes are two distinct memory states, working or short term memory, and long term memory. Working memory is roughly analogous to awareness and is the memory state containing the info rmation that is currently in process. Long term memory, on the other hand, is roughly analogous to a computer hard drive holding a great deal of information that is stored in an organized fashion (Farquhar & Surry, 1995). To be retained in long term memor y, information must first be processed by working memory. However, working memory is limited in both its capacity and duration, which can hinder the learning process (Cooper, 1998). Unlike working memory, long term memory holds information for extended pe riods of time (Farquhar & Surry, 1995). According to Mayer (1992), learning occurs with the selection and organization of relevant content, and the integration of the new content with prior

PAGE 63

53 knowledge. Working memory plays a key, albeit limiting, role in t he selection, organization, and integration of the new content. Working memory is the pathway between long term memory and the outside world and, as such, is the place where the instructional message and prior knowledge are integrated (Tabbers, 1999). Wo rking memory, therefore, is at the center of the learning process. Because working memory is limited (Miller, 1956), high cognitive load or strain on working memory will inevitably impede the learning process. Miller described the limits of working memor y as the number of chunks (seven plus or minus two) of processing capacity that working memory possessed. The chunks however could contain variable amounts of information and could be influenced by several perceptual dimensions. As a result, the additi on of elements such as color, size, or spatial location can increase the processing capacity of working memory. The goal of educational theories relating to cognition and instruction is the generation of useful instructional techniques (Sweller & Chandle r, 1991). Cognitive load theory describes the information processing system. The essential tenet of cognitive load theory is that learning outcomes can be improved through greater understanding of the role of working memory in the learning process (Cooper 1998). Better understanding of the data and the usefulness of its predictive power are key elements of theories in education, but unless these elements are followed by improved teaching methods and then the ultimate goal of education theory is missed (S weller & Chandler, 1991).

PAGE 64

54 In the field of education, instructional message design is used to create environments where cognitive load is reduced and learning is enabled. Evaluating web sites according to instructional message design principles is a method of appraising web site effectiveness, but in the commercial sector such techniques have not typically been used. Complex instructional messages can impose on working memory by presenting multiple stimuli resulting in a sensory overload that impedes proce ssing from working memory to long term memory. When impositions on working memory are reduced, a greater degree of cognitive engagement results. There are various instructional message design strategies designed to increase cognitive engagement through re duction on impositions on working memory (Farquhar & Surry, 1995). Use of supportive message design elements, such as effective text and graphics, increases attention and motivation, which, in turn, reduces cognitive load. Reduced cognitive load enhances the processing of the intended message into long term memory. Figure 1 2 illustrates these relationships. Application of Cognitive Load Theory and Instructional Message Design Principles to Tourism Information Search CVBs have, as their central missio n, an obligation to present information about destinations they represent in a manner that facilitates information processing by the consumer. This warrants consideration of the role o f supportive message design in the reduction of cognitive load and, thu s, in the facilitation of information processing for tourists. Figure 2 1 represents how

PAGE 65

55 supportive message design fits within a consumer information acquisition and processing model. Vogt and Fesenmaier (1998) characterized information processing as a cr itical link in the model. When information is acquired via the Internet, supportive message design should reduce cognitive load of the user, thereby facilitating efficient information processing. Figure 2 1. The role of supportive message design in the facilitation of consumer information acquisition and processing. The process of information search and evaluation involves learning and learning styles (Hodkinson & Kiel, 2003). The primary responsibility of an informational web site is to meet the expectations and needs of visitors by creating an environment where lea rning can occur (Hubbard, 1998). When used as an instructional technology, therefore, web pages must be designed to provide an environment that can enable a successful learning experience.

PAGE 66

56 Instructional web sites can be made more effective through utilization of the principles offered through the study of instructional message design. The focus of instructional message design is linking learning theory and educational practice by prescribing learning conditions that are likely to result in desired outcomes (Fleming & Levie 1993). Effective message design should enhance information acquisition and processing by reducing co gnitive load. Design elements can enhance or inhibit both motivation and attention. Sound message design serves to maximize motivation and attention, while eliminating elements that detract from these goals. The most critical elements of instructional me ssage design are the text and graphics elements used to convey effective messages. Both text and graphics can be used as motivational tools and as methods to get and maintain a learners attention. The appropriate use of textual features can increase atten tion and influence motivation to learn (Keller & Burkman, 1993). According to cognitive load theory, the limitation of working memory capacity is a key issue in the learning process. For web designers, content must be presented in such a way that the use rs working memory is focused on relevant information and is not taxed by irrelevant or distracting features of instructional material. The value of instructional message design is its ability to decrease extraneous cognitive load and direct working memor y resources to the learning process itself by limiting the effect of the presentation features of instructional material (Tabbers, 1999).

PAGE 67

57 By providing information th rough supportive message design, CVBs can enhance the effectiveness of their destination me ssage. This study examines use of supportive message design elements on CVB web sites. Chapter 3 outlines the methods used in the study.

PAGE 68

58 CHAPTER 3 METHODS The purpose of this study was to investigate the viability of travel and tourism web site s as instructional environments, and to determine whether geographic differences exist with respect to web site utilization of instructional message design strategies on those web sites. That is, this study examine d whether and how effective instructio nal message design is used by convention and visitors bureaus to deliver destination message via the Internet. Specifically, do CVBs utilize the tenets of instructional message design? This chapter includes a description of sampling procedures and the pop ulation of web sites. Succeeding sections of this chapter include details of the research design and treatment of the data. The results of this study can be used to evaluate web site s in the context of information reten tion and assess the viability of travel and tourism web sites as instructional environments This study was conducted over an eight week period during the spring and summer of 2006. The study was designed to answer the following research questions: 1. What is the prevalence of message design elements on CVB web sites? 2. Which message desig n elements are CVB web sites using effectively to deliver destination messages? 3. How many underlying instructional message design conditions that enable the desired retention outcomes are present on CVB web sites?

PAGE 69

59 4. Do regional or geographic differences exi st in the use of message design elements on CVB web sites? 5. What web page elements do web design professionals believe to be most important for destination message delivery? The following sections provide details about the quantitat ive and qualitative research methods that were used to carry out this study. Methods: Research Questions 1 4 The purpose of Research Questions 1 4 was to examine the use of message design elements on CVB web sites. The following sections outline the proc edures employed to answer these questions. Item Review The initial pool of message design items was gathered from a literature review of published articles, from both popular and scientific sources, related to text and graphics in instructional message des ign. A list of 93 elements was compiled. A review of the professional literature provided external validity for 70 of the message design elements and the list was subsequently reduced to that number. Further review of the human computer interaction lit erature provided external validation from studies ( similar to message design ) conducted using computer screen (i.e., not print) media. The resultant item list was reviewed by two design experts: Dr. Lee Mullally (instructional message design) and Dr. Benja min Lok (human computer interaction). The experts provided input to improve each items clarity, readability, and appropriateness, and to recommend removal of items deemed redundant. A final list of 46 variables (Appendix A) was used in this study.

PAGE 70

60 Sample CVB web sites were chosen as the unit of analysis due to th eir role as providers of information for the tourism industry. A population (N=1 499) of CVB web sites was identified from a listing of worldwide sites stratified into nine geographic regions and on the web site TourismBureaus.com (http://www.tourismburea us.com/default.asp). The final sample was randomly selected from each of nine world regions (Table 3 1 ). Using a power analysis to determine a minimum N and adding a sufficient number to allow for attrition, a sample of 625 CVB web sites (from a populatio n of 1,499) was randomly selected for data collection. Due to the transient nature of web sites, some of the Table 3 1. Sample from CVB world regions Number of listed CVBs per geographic region (total = 1499) Sample N for region % of Region Sampled Conf idence interval Asia (42) 30 71 4.22 Africa (20) 18 90 3.27 Central/South America (30) 25 83 3.55 Canada (77) 55 71 3.10 Europe (153) 109 71 2.20 Caribbean (41) 30 73 4.09 Middle East (15) 13 87 4.48 United States (1082) 277 26 2.21 Oceani a/South Pacific (39) 31 80 3.52 Total 588 links were not functional at the time of data collection. If another link from the same geographic strata was available, a missing link was replaced with another randomly selected site. Using this procedure the resulting final stratified random

PAGE 71

61 sample size was 588. This sample was examined to answer research questions 1 through 4, whereas research question 5 was answered through a Delphi study. The final sample of 588 CVB web sites was examined using quantitative c ontent analysis, a method well suited to the study of communication (Babbie, 2001), to assess the number and types of instructional message design elements utilized. Data Collection P rocedures The source of content assessed from each web site was limited to the initial information page on each web site to standardize data collection. The CVB web sites were visited using the web browser Mozilla Firefox 1. 5 for Windows The computers used i n this study were set at default preferences for display properties, using 32 bit color and a monitor resolution of 1024x768 (OneStat, 2005). Web sites were assayed for content using a checklist type coding sheet (Appendix A). During data collection, eac h element was assessed on a Likert type scale ranging from 1 to 4, where 1 = poor, 2 = fair, 3 = good, and 4 = excellent. If any of the listed items were not applicable to the web page under scrutiny, the N/A box was checked. Reliability The reliabilit y of the scale items was ascertained using SPSS to indicate the precision of the measurement, or assay how large the errors of measurement were likely to be. The reliability coefficient Cronbachs Alpha ( a measure of internal consistency based on the ave rage inter item correlation ) ranges between 0 and 1, where 1 denotes 100% error free measurement.

PAGE 72

62 Interrater R eliability Th re e raters conducted the evaluation of the web sites. T he researcher served as one of the raters. The two other rater s were adults who (a) were experienced with computer use, (b) had backgrounds in business, and (c) participated in a two hour training session conducted by the researcher A random sample of 200 sites was assessed by all raters to establish interrater reliability. Stemler (2004) avers that the practice of depicting interrater reliability as a single, unified concept is misleading, and that statistical methods for computing int errater reliability can be more accurately classified into three categories: consensus estimates, measurement estimates, and consistency estimates. Researchers are cautioned that concomitant with different approaches to estimating interrater reliability ar e implications for how ratings across multiple judges should be summarized. Consensus estimates of interrater reliability, such as percent agreement, assume that the data reported by reasonable observers should reflect exact agreement about how to apply a scoring rubric to observed data (Stemler, 2004). Consequently, if raters can come to exact agreement on how to apply the rating scale, then the raters may be assumed to share a common interpretation of the scoring rubric. Percent agreement has some distin ct disadvantages, however. Methods for establishing interrater reliability by using percentage agreement between raters fail to account for the role of chance in recording data. In content

PAGE 73

63 analysis, some chance is bound to occur, with resultant inflation of agreement percentages, especially where the number of raters is low (MacLennan, 1993). A further disadvantage of determining interrater reliability by percentage agreement is that the amount of time to train raters to come to exact agreement can be sig nificant, and may be impossible if the number of raters is large. Training raters to the point of forced consen sus may actually reduce the objective independence of the ratings, and threaten the validity of the study (Linacre, 2002). Finally, consensus est imates such as percentage agreement can be overly conservative if two raters exhibit systematic differences in scoring items, but simply cannot be trained to reach an exact consensus (Stemler, 2004). Measurement estimates (principle components factor anal ysis, facet rater indices and fit statistics) assume that researchers use all of the data available from all raters, including inconsistent ratings, when computing summary scores, as it is the accumulation of information, not the ratings themselves, that i s desirable. When using measurement estimates, it is not necessary for raters to come to a consensus on application of a scoring rubric because differences in ratings can be estimated in the production of each raters final score (Stemler, 2004). Consisten cy estimates of interrater reliability, such as Pearsons correlation coefficient, Spearmans rank coefficient and C ronbachs alpha are based on the assumption that each rater is consistent in classifying the observed data according the raters interpretat ion of the scoring rubric. Pearsons coefficient requires assay of one pair of raters and one item at a time, while

PAGE 74

64 Spearmans coefficients requires all raters to rate all cases; as such, neither technique was applicable to this data set. Crocker and Algin a (1986) recommended that in studies where multiple raters are used, a high Cronbachs alpha would produce a consistency estimate of interrater reliability demonstrating that the bulk of the variance in the total composite score was true score variance and not error variance. An advantage of using Cronbachs alpha is its ability to produce a single estimate of interrater reliability across multiple judges (Stemler, 2004). A C ronbachs alpha score above 0.7 is acceptable for demonstrating an acceptable esti mate of interrater reliability (Barrett, 2001). In this study, interrater reliability was established by three methods. Initially, an analysis of variance on all items using the selection variable rater identified two items that exhibited significant dif ferences between raters. The variables 56K modem and ISDN 128 were removed from further analysis due to unreliability. Data recorded on Likert type items were assayed using principal components factor analysis. This method can be used as a measurement estimate of interrater reliability by providing a single statistic, permitting direct comparison of the raters, and producing an empirical estimate of the extent to which raters applied the rating scale across variables. If the shared variance is high (gre ater than 60%), then this technique indicates that the raters are rating a common construct (Stemler, 2004). Each rater scored above the threshold level of 60% (Table 3 2 ).

PAGE 75

65 Table 3 2 Rater summary scores on primary component Rater Percent of Varianc e Explained 1 68.71 2 67.40 3 70.23 Finally, C ronbachs alpha was used to produce a consistency estimate (Crocker & Algina, 1986) of interrater reliability support ing the supposition that the variance in the total composite score was true sco re variance and not error variance The C ronbachs alpha for the Likert type items was .86, indicating a high degree of interrater reliability. Other Variables Measured In addition to the message design elements addressed by the rating scale, on each CVB web site evaluated, several other variables were measured as well : download time, readability, and legibility. Each of these elements contributes to the user s experience while visiting a web site. Download time Download time for each CVB site was measured by the Web Page Analyzer v0.961, an online tool provided from Web Site Optimization (http://www.websiteoptimization.com/services/analyze /). The URL for each site was entered to determine the download speed for various Internet connections. Readability Readability of CVB web sites was determined using an online measurement tool provided through the web developer resource site JuicyStu dio.com (http://juicystudio.com/services/readability.php#readresults), where URLs of selected web sites can be submitted for web page assessment by

PAGE 76

66 three reading level algorithms. Although reading level algorithms only provide a rough guide, the Gunning Fo g, Flesch Reading Ease, and Flesch Kincaid readability tests can be useful in determining how readable web site content is, and can give some indication as to whether message designers have developed content at the reading level of the intended audience. The Gunning Fog index determines how many years of schooling would be required for a reader to understand the web site content, where the lower the number, the more understandable the content should be (scores over 17 are reported as 17 to post graduate le vel). Results for the Flesch Reading Ease test are reported on a 100 point scale, where the lower the score, the more difficult the content is to understand. Content writers are encouraged to develop text that scores in the 60 to 70 range. The Flesch Kin caid grade level, like the Gunning Fog index, is a measure of the number of years of schooling that would be required for a reader to understand the text (negative scores are reported as zero, and scores over 12 are reported as 12). Legibility The on screen legibility or distinctness of the text and headings was assessed by software provided through the Web Accessibility Tools Consortium (http://www.wat c.org). The Color Contrast Analyzer v1.0 uses a color contrast algorithm to test whether foreground/ background color and brightness combinations are greater than a set range, providing acceptable screen visibility for both normal vision and three types of vision impairments. The Color Contrast Analyzer employs formulas suggested by the World Wide Web Con sortium (http://www.w3.org/TR/AERT#color contrast) to determine

PAGE 77

67 color brightness ((red value X 299) + (green value X 587) + (blue value X 114)) / 1000 ) and color difference (maximum (red value 1, Red value 2) minimum (red value 1, red value 2)) + (maximu m (green value 1, green value 2) minimum (green value 1, green value 2)) + (maximum (blue value 1, blue value 2) minimum (blue value 1, blue value 2)) where color brightness must be greater than 125 and color difference must be greater than 500. The Color Contrast Analyzer also tests color brightness/difference for 3 types of color vision impairments (i.e., protanopia, deuteranopia, and tritanopia) and indicates whether they meet threshold requirements. Treatment of the Data Data were analyzed using SPSS (Statistical Package for the Social Sciences) v.12.0 for Windows. Item frequencies, means and percentages were recorded to compare the descriptive statistics of the sample Missing values for Likert type items were replaced using the SPSS function Replace Missing Values by selecting the linear trend at point method. Linear trend at point r eplaces missing values by regression on an index variable where missing values are replaced with their predicted values. The linear trend at point method maintains a degree of overall variance within variables, where simple mean substituti on would not. Factor analysis a method of data reduction that attempts to identify underlying factors that explain the intercorrelations within a set of variables is used to identify a limited number of factors that explain most of the variance observed in a much larger number of variables. Factor analysis of the Likert type

PAGE 78

68 items was performed to determine whether any sub scales of the construct exist within the data collected. The reliability of any sub scales was individually assessed. Those subscale s with acceptably high reliability (Cronbachs alphas .7) were selected for inclusion in the final index. The result of this analysis yielded a multi dimensional model of a web based learning environment. Methods: Research Question 5 To allow for mor e meaningful interpretation of the results of this study and to understand its implications for practice, it was important to ascertain which web page elements web design professionals believe to be most important for destination message delivery. The con tinually and rapidly changing nature of the Internet and of web design practices required interpretation of results via consultation with practicing professionals, rather than relying on published research alone. Therefore, research question 5 was address ed through analysis of a Delphi study conducted on a panel of web design professionals. The Delphi method is a systematic rigorous and effective method that can produce potent and valid answers to research questions (Clayton, 1997). The objective of the Delphi method is to obtain a reliable consensus of opinion of a selected panel of experts through a series of detailed questionnaires developed through ongoing feedback. The Delphi method uses multiple iterations with researcher controlled feedback (Murr y & Hammons, 1995) from the individual experts and avoids direct interaction of the individual experts with one another to control unnecessary bias. The initial questions are intended to elicit the respondents reasoning that went

PAGE 79

69 into their replies, facto rs considered to be relevant to the study, the respondents judgment of these factors, and other information the respondents feel may be helpful in reevaluating various factors that may provide more illuminating answers to the primary question (Dalkey & H elmer, 1963). The Delphi method of controlled interaction among panel members provides an advantage over more conventional uses of experts, such as roundtable discussions, by encouraging independent thought and aiding the gradual formation of considered o pinion. Linstone and Turoff (2001) offer a number of situations where the Delphi technique could be successfully utilized, including situations where understanding of the topic can benefit from subjective judgments on a collective basis, and where time a nd cost make group meetings inefficient. Computer based Delphi studies enable respondents to participate in an asynchronous manner (Turoff et al., 2004). A perceived weakness of the traditional Delphi technique was the relatively long periods of time requ ired to administer the method by mail. An e Delphi system speeds the process of data collection and feedback and limits data loss from nonparticipation (Chou, 2002). Delphi P anel S election In the Delphi method, group consensus is the goal of the proj ect. The group is comprised of a panel of experts. According to Scheele (1975) experts should be selected from among stakeholders (those who may be directly affected by the topic under study), experts (those who have special experience or education), and facilitators. The selection of experts (n=4) for this study followed the following criteria: e ach expert possessed a special knowledge of web site

PAGE 80

70 creation, including selection of text and graphical content ; e ach expert has been employed by the tourism industry and charged with web design ; and e ach expert professed a keen interest in the topic under study First R ound: Questionnaire The first round checklist questionnaire (Appendix B) was distributed to respondents along with detailed guidelines on its completion. The questionnaire was designed to investigate the importance ratings panel members assign to various web site elements. The panel was advised as to the depth of responses sought, and given instruction for ratin g the items in the study so that all respondents interpreted the design items in a similar fashion. A four point Likert type response scale was adopted for its familiarity, and to easily identify areas of agreement and disagreement. The first round in th e Delphi study was designed to illuminate the respondents initial positions on the subject at hand. With respect to the topics to be evaluated, the first round questionnaire identified which items from the questionnaire the respondents agree upon, and wh ich items provoked disagreements. A complete record of the first round questionnaire responses was assembled, and analysis of this record identified agreements and disagreements among panel members on every item under consideration. Second Round : S ubject Exploration The second round was predicated on the responses and analysis of the first round questionnaire. The Delphi process stops when consensus among the respondents has been reached or when saturation has occurred and no new

PAGE 81

71 info rmation is forth coming Subsequent rounds provide the panel with an opportunity to evaluate both their own responses and group answers from the previous round. In the second round, respondents were asked to consider their answers from the initial round in light of the group responses, and to justify and defend their positions. In the second round, respondents began to explore why they believed some web site elements to be more important than others and the underlying reasons guiding perceptions of rela tive importance. In the second round of data analysis, the researcher synthesized new data based on items of group consensus, and by identifying those items where differences of opinion exist ed Third R ound: Consensus In the third round, the panel r eviewed the remaining areas of dissention and compared individual views with group views. At this point, most it ems had been explored and reevaluated. Analysis of the third round of responses illuminated some general agreement among the respondents with respect to the research question given to them, and the reasons for their positions.

PAGE 82

72 CHAPTER 4 RESULTS Before addressing the research questions, i tem analysis was performed following data collection to assess the measurement properties of individual items. Since a Likert type scale measures between ind ividual differences, recorded values must fall at different levels of the scale or the items would not indicate any information concerning individual differences on the variable in question, and thus yield nothing to the overall measurement of the construc t. Each item was analyzed for response variability, response location, and response discrimination. Response variability. Penfields V was used to evaluate variability in item responses (the index of variation ( ) could have been used, but is more af fected than V by items having low response rates). For nearly all variables, Penfields V was in excess of .90, indicating that the items were doing a good job of measuring real differences on different facets (Table 4 1) One variable, legible font, ha d medium variability (.75 # V # .90), and could be considered for revision. Three items, running text in lower case letters, avoid jargon/slang, and familiar font exhibited poor variability and became candidates for removal from further analysis. Res ponse location. Items were further assayed for response location (Table 4 2). Response location indicates whether the majority of individuals lie at

PAGE 83

73 Table 4 1 Response variability of the Likert type items Variable N Standard Deviation Penfields V Re sponse Variability headings attract attention 588 0.81 1.08 Good headings identify subject matter 588 0.97 1.29 Good loss text overview 588 1.10 1.46 Good whitespace use 588 1.06 1.41 Good text distinct segments 588 .90 1.2 Good short blocks te xt 588 0.93 1.25 Good running text lower case 588 0.46 0.61 Poor lines text 8 to 10 words 588 1.04 1.39 Good typographic queuing 588 1.00 1.33 Good embedded hotlinks for key terms 588 1.20 1.60 Good highlighted key words 588 0.77 1.03 Good lead ing (interlinear spacing) 588 0.93 1.24 Good column width 588 1.01 1.35 Good avoid jargon, slang 588 0.52 0.69 Poor same info different formats 588 1.11 1.48 Good well organized text 588 0.96 1.28 Good explicit text 588 0.92 1.23 Good content g eneralized to specific 588 1.26 1.67 Good familiar font 588 0.40 0.53 Poor legible font 588 0.66 0.88 Medium font size appropriate 588 0.82 1.09 Good multiple type styles 588 0.84 1.11 Good captions support graphics 588 1.06 1.41 Good captions near to corresponding graphics 588 1.08 1.44 Good graphic overview scrolling 588 1.10 1.47 Good graphics gain attention 588 1.06 1.42 Good favorable first impression 588 1.01 1.35 Good picture depiction accurate 588 1.00 1.33 Good picture is aest hetically satisfactory 588 1.10 1.46 Good picture is technically acceptable 588 0.90 1.19 Good picture fits in area 588 0.99 1.32 Good picture fits with other pictures 588 0.92 1.23 Good picture overload 588 1.02 1.36 Good graphics support textua l content 588 1.05 1.40 Good consistent color scheme 588 0.72 0.96 Good realistic colors used in graphics 588 0.81 1.08 Good color media preferable to B&W 588 0.89 1.19 Good

PAGE 84

74 Table 4 2. Response location of the survey items Variable Ideal Location Mean Absolute Difference Response Location headings attract attention 2.5 2.42 0.08 Good headings identify subject matter 2.5 3.03 0.53 Medium loss text overview 2.5 2.79 0.29 Good whitespace use 2.5 3.08 0.58 Medium text distinct segments 2.5 3.54 1 .04 Poor short blocks text 2.5 3.42 0.92 Medium running text lower case 2.5 3.88 1.38 Poor lines text 8 to 10 words 2.5 3.06 0.56 Medium typographic queuing 2.5 2.23 0.27 Good embedded hotlinks for key terms 2.5 2.10 0.40 Good highlighted key words 2 .5 1.37 1.13 Poor leading (interlinear spacing) 2.5 3.46 0.96 Medium column width 2.5 3.21 0.71 Medium avoid jargon, slang 2.5 3.80 1.30 Poor same info different formats 2.5 2.20 0.30 Good well organized text 2.5 3.29 0.79 Medium explicit text 2.5 3. 20 0.70 Medium content generalized to specific 2.5 1.99 0.51 Medium familiar font 2.5 3.89 1.39 Poor legible font 2.5 3.70 1.20 Poor font size appropriate 2.5 3.20 0.70 Medium multiple type styles 2.5 3.30 0.80 Medium captions support graphics 2.5 1. 70 0.80 Medium captions near to corresponding graphics 2.5 1.71 0.79 Medium graphic overview scrolling 2.5 3.02 0.52 Medium graphics gain attention 2.5 2.52 0.02 Good favorable first impression 2.5 2.55 0.05 Good picture depiction accurate 2.5 3.16 0. 66 Medium picture is aesthetically satisfactory 2.5 2.88 0.38 Good picture is technically acceptable 2.5 3.32 0.82 Medium picture fits in area 2.5 3.30 0.80 Medium picture fits with other pictures 2.5 3.21 0.71 Medium picture overload 2.5 3.38 0.88 Me dium graphics support textual content 2.5 2.52 0.02 Good consistent color scheme 2.5 3.64 1.14 Poor realistic colors used in graphics 2.5 3.63 1.13 Poor color media preferable to B&W 2.5 3.61 1.11 Poor

PAGE 85

75 the high, medium, or low end of the item scale. I t is preferable to have the mean of the item response distribution lie near the center of the item scal e. Wi th a scale ranging from 1 to 4, it is preferable to have a mean near 2.5. Nine items exhibited poor response location became candidates for removal from analysis. Item discrimination. Item discrimination refers to how well the item distinguishes between high and low levels of the construct being measured. Item discrimination was ascertained by finding the item total correlation index. The item tot al correlation measures the correlation between the individual item scores and the scale score. For a Likert type scale, an item discrimination score .5 is desirable, whereas # .2 is unacceptable. Results of the item discrimination analysis are present ed in Table 4 3. Six of the items have poor item discrimination, but in no case would the alpha appreciably rise if the item were deleted. Table 4 4 presents a composite of the item analysis findings and a determination of which items would be rejected f or further analysis. Thirty two of the variables had acceptable composite scores for response variability, location, and item discrimination and appeared to be acceptable measures of the construct. In some cases (e.g., text in distinct segments), items ha d poor ratings on one of the three facets considered in item analysis. Retention of these items was a judgment call and, because none of the overall ratings were sufficiently low, the items were not excluded. For five variables, running text lower case, highlighted key words, avoid jargon, slang, familiar font, and legible font, overall item analysis scores were sufficiently low to identify those items as poor measures of the construct. These items were

PAGE 86

76 Table 4 3. Item discrimination of Likert type items (alpha=.86). Variable Item Total Correlation Alpha, if deleted Item Discrimination headings attract attention .133 .863 Poor headings identify subject matter .283 .860 Medium loss text overview .174 .863 Poor whitespace use .457 .856 Medium text distinct segments .482 .856 Medium short blocks text .398 .857 Medium running text lower case .356 .859 Medium lines text 8 to 10 words .316 .859 Medium typographic queuing .272 .860 Medium embedded hotlinks for key terms .222 .862 Medium high lighted key words .062 .864 Poor leading (interlinear spacing) .510 .855 Good column width .447 .856 Medium avoid jargon, slang .066 .863 Poor same info different formats .447 .856 Medium well organized text .512 .855 Good explicit text .462 .856 Med ium content generalized to specific .127 .865 Poor familiar font .208 .861 Medium legible font .237 .861 Medium font size appropriate .324 .859 Medium multiple type styles .303 .859 Medium captions support graphics .306 .860 Medium captions near to corresponding graphics .287 .860 Medium graphic overview scrolling .231 .862 Medium graphics gain attention .532 .854 Good favorable first impression .629 .852 Good picture depiction accurate .510 .855 Good picture is aesthetically satisfactory .456 856 Medium picture is technically acceptable .505 .855 Good picture fits in area .466 .856 Medium picture fits with other pictures .578 .853 Good picture overload .242 .861 Medium graphics support textual content .349 .859 Medium consistent color sch eme .419 .857 Medium realistic colors used in graphics .411 .857 Medium color media preferable to B&W .412 .857 Medium

PAGE 87

77 Table 4 4. Item analysis composite. Variable Response Variability Response Location Item Discrimination Item Utility headings attra ct attention Good Good Poor Accept headings identify subject matter Good Medium Medium Accept loss text overview Good Good Poor Accept whitespace use Good Medium Medium Accept text distinct segments Good Poor Medium Accept short blocks text Good Mediu m Medium Accept running text lower case Poor Poor Medium Reject lines text 8 to 10 words Good Medium Medium Accept typographic queuing Good Good Medium Accept embedded hotlinks for key terms Good Good Medium Accept highlighted key words Good Poor Poor Reject leading (interlinear spacing) Good Medium Good Accept column width Good Medium Medium Accept avoid jargon, slang Poor Poor Poor Reject same info different formats Good Good Medium Accept well organized text Good Medium Good Accept explicit te xt Good Medium Medium Accept content generalized to specific Good Medium Poor Accept familiar font Poor Poor Medium Reject legible font Medium Poor Medium Reject font size appropriate Good Medium Medium Accept multiple type styles Good Medium Medium A ccept captions support graphics Good Medium Medium Accept captions near correspond. graphics Good Medium Medium Accept graphic overview scrolling Good Medium Medium Accept graphics gain attention Good Good Good Accept favorable first impression Good G ood Good Accept picture depiction accurate Good Medium Good Accept picture is aesthetically satisfactory Good Good Medium Accept picture is technically acceptable Good Medium Good Accept picture fits in area Good Medium Medium Accept picture fits with other pictures Good Medium Good Accept picture overload Good Medium Medium Accept graphics support textual content Good Good Medium Accept consistent color scheme Good Poor Medium Accept realistic colors used in graphics Good Poor Medium Accept color media preferable to B&W Good Poor Medium Accept

PAGE 88

78 removed from further scale analysis, leaving a final set of 32 Likert type items remaining Following the response variability, response location and item discrimination procedures, items were considered f or removal from further analysis based upon their relative scores. Each of the items that had acceptable response variability, location and item discrimination were retained, as they appeared to be good measures of the construct. Six variables were underrepresented in the sample. Although still useful for descriptive statistics, these six variables occurred too infrequently on CVB web sites for meaningful statistical analysis. These six items were: (1) abbreviations, acronyms a nd initialisms, (2) metaphors make unfamiliar familiar, (3) graphics easy to read, (4) meanings of icons immediately obvious, (5) information w/color also w/out color, and (6) graphics used when messages are complex. Results of Data Analysis Following the item analysis, data were further analyzed to answer each research question. The following sections pres ent the results of these analyses. Research Question 1 To answer Research Question 1 What is the prevalence of message design elements on CVB web sites ? descriptive statistics were compiled to reflect the appearance of message design elements on the 588 selected CVB web sites (Table 4 5 ). Instructional message design elements, in general, app eared throughout the sampled web sites. A few elements, such as

PAGE 89

79 Table 4 5 Prevalence of instructional message design elements on CVB web sites. Message Design Ele ment Number of Sites Percent of Sites white space use 585 99.5% loss text overview 584 99.3% font size a ppropriate 583 99.1% multiple type styles 583 99.1% familiar font 582 99.0% legible font 581 98.8% consistent color scheme 581 98.8% head ings attract attention 580 98.6% favorable first impression 579 98.5% headings identify subject matter 578 98.3% leading/interlinearspacing 575 97.8% column width 574 97.6% well organized text 574 97.6% running text lower case 573 97.4% expli cit text 572 97.3% color media preferable to b lack & w hite 572 97.3% realistic colors used in graphics 570 96.9% lines text 8 to 10 words 569 96.8% text distinct segments 567 96.4% short blocks text 567 96.4% picture overload 567 96.4% graphi c overview scrolling 565 96.1% graphics gain attention 565 96.1% picture is aesthetically satisfactory 564 95.9% picture is technically acceptable 564 95.9% picture fits in area 562 95.6% picture depiction accurate 559 95.1% content generalized to specific 555 94.4% same info different formats 518 88.1% picture fits with other pictures 513 87.2% macrosignals under italics bold 498 84.7% typographic queuing 497 84.5% imbedded hotlinks for key terms 495 84.2% captions support graphics 487 82.8% captions placed near to corresponding graphics 485 82.5%

PAGE 90

80 Table 4 5. Continued. Message Design Element Number of Sites Percent of Sites graphics support textual content 472 80.3% highlighted key words 471 80.1% avoid jargon, slang 44 8 76.2% meanings of icons immediately obvious 293 49.8% abbreviations, acronyms and initialisms 244 41.5% graphics easy to read 227 38.6% information w/color also w/out color 143 24.3% graphics used when messages are complex 138 23.5% metaphors make unfamiliar familiar 41 7.0% typographic queuing embedded hotlinks captions support graphics captions placed near to corresponding graphics graphics support textual content highlighted key words avoid jargon, slang meanings of icons immediately obvious abbreviations, acronyms and initialisms and graphics easy to r ead appeared on fewer than 500 web sites, while Information w/color also w/out color graphics used when messages are complex appeared on fewer than 200 visited sites. The message design element metaphors make unfamiliar familiar was used on only 41 (7%) of 588 sites. Research Question 2 Research Question 2 Which message design elements are CVB web sites using effectively to deliver destination messages ? w as also answered using descriptive statistics Effectiveness of design element use was operationalized by rat ing each item on each CVB web site using a four point Likert type scale with 1 = poor, 2 = fair, 3 = good, a nd 4 = excellent The mean rating scores and standard deviations are reported in Table 4 6

PAGE 91

81 Table 4 6. Ratings of message design element use on CVB web sites. Message Design Element Number of Si tes Mean Score St andar d Dev iation familiar font 582 3.89 0.40 running text lower case 573 3.88 0.47 avoid jargon, slang 448 3.80 0.60 abbreviations, acronyms and initialisms 244 3.77 0.63 legible font 581 3.70 0.66 consistent color scheme 581 3.64 0. 73 realistic colors used in graphics 570 3.62 0.82 color media preferable to black & white 572 3.61 0.90 text distinct segments 567 3.47 0.92 leading/interlinearspacing 575 3.46 0.94 short blocks text 567 3.42 0.95 picture overload 567 3.39 1.04 met aphors make unfamiliar familiar 41 3.34 1.06 picture is technically acceptable 564 3.32 0.91 multiple type styles 583 3.30 0.84 well organized text 574 3.29 0.97 picture fits in area 562 3.29 1.01 column width 574 3.21 1.02 font size appropriate 583 3.20 0.82 explicit text 572 3.20 0.93 picture fits with other pictures 513 3.20 0.98 picture depiction accurate 559 3.15 1.02 white space use 585 3.08 1.06 lines text 8 to 10 words 569 3.06 1.06 headings identify subject matter 578 3.03 0.98 graphic overview scrolling 565 3.02 1.12 picture is aesthetically satisfactory 564 2.85 1.07 loss text overview 584 2.79 1.10 meanings of icons immediately obvious 293 2.58 1.32 favorable first impression 579 2.55 1.02 graphics gain attention 565 2.51 1.08 graphics support textual content 472 2.50 1.16 headings attract attention 580 2.37 0.82 Information w/color also w/out color 143 2.27 1.21 graphics easy to read 227 2.25 1.21 typographic queuing 497 2.23 1.09

PAGE 92

82 Table 4 6. Continued Message Design El ement Number of Sites Mean Score Standard Deviation macrosignals under line, italics bold 498 2.21 1.09 same info different formats 518 2.20 1.18 imbedded hotlinks for key terms 495 2.09 1.31 content generalized to specific 555 1.99 1.29 captions sup port graphics 487 1.70 1.16 captions placed near to corresponding graphics 485 1.70 1.19 graphics used when messages are complex 138 1.58 1.09 highlighted key words 471 1.37 0.87 Research Question 3 Research question 3 How many underlying instructiona l message design conditions that enable desired learning outcomes are present on CVB web sites? was answered through exploratory factor analysis of the message design variables to illuminate the conditions, or underlying factors, that explain the shared va riance of a total number of Likert type variables. A p rincipal c omponent a nalysis using v arimax rotation with Kaiser n ormalization was conducted and the results are presented in Table 4 7 Nine factors explaining 64.7% of the varia nce were illuminated by the factor analysis The item headings attract attention did not load on any factor and was removed. One of the factors was a single item factor (content general to specific ) and was removed from further factor analysis, leaving eight factors that explained 61.4 % of the variance. The default setting for factor retention in most statistical software is to retain all factors with eigenvalues over 1.0, although that may be the least desirable method for determining the number of fa ctors to retain (Costello & Osborne, 2005). A scree test, widely utilized as a better method to determine

PAGE 93

83 factor retention, is an examination of the plot of eigenvalues to determine the natural break point where the curve flattens. The number of data poin ts above the break is recommended as the number of factors to retain (Costello & Osborne). A scr ee plot of the factor loadings for this data set suggested retention of four dimensions in the final model (Figure 4 1). Figure 4 1. Scree plot of factor l oadings. Based on the scree test, four f actors were selected for retention. Factor 1 was graphic presentation which consisted of eight items and explained 21.7% of the variance. Factor 2 was text presentation consisting of 5 items and explaining 11.7% of the variance. Factors 3 and 4 each cons isted of two items. In exploratory factor analysis, the decision to retain factors with as few as two items may be appropriate (StatSoft, 2006). In this study, both factor 3 and factor 4

PAGE 94

84 contained highly correlated items with strong factor loadings, which subsequently proved to be highly reliable measures, as well (Table 4 7). That evidence, taken together with the scree test recommendation of four factors retained, justified keeping factor 3 ( visual text ) and factor 4 ( graphic/text support) in the final model. The factor visual text explained 6.48% of the variance, while the factor graphic/text support explained 4.04% of the variance The reliability of the scale was ascertained to indicate the precision of the measurement or more specifically, to assay how large the errors of measurement were likely to be. The reliability coefficient (alpha) ranges between 0 and 1, where 1 denotes 100% error free measurement. Table 4 7 presents the reliability of the sub scales. The four fac tors selected for retention each exhibited acceptably high reliability. Graphic presentation (alpha = .89), text presentation (alpha = .75), visual text (alpha = .84) and graphic/text support (alpha = .97) each proved to be reliable measures. The remaining four factors each had a Cronbach alpha of less than .7 and cannot be considered reliable measures. The item short blocks of text loaded on both factor 2 and factor 3. Short blocks of text is conceptually more aligned with the factor 2 attributes of organ ization and segmentation than the factor 3 attributes of line length/column width. Additionally, retaining short blocks of text on factor 2 raised the alpha from .70 to .75, while removing short blocks of text from factor 3 raised the factor 3 alpha from .75 to .84, indicating greater reliability in each case. Retention or exclusion of items in exploratory factor analysis requires judgment to produce the cleanest fit of factors (Costello & Osborne, 2005). In this case, the conceptual

PAGE 95

85 Table 4 7 Factor a nalysis and reliability of the message design variables. Message Design Domain Factor Loading Eigen value % of Variance Cron bachs alpha Alpha if item deleted Mean Std Dev Factor 1 Graphic Presentation 6.94 21.70 .89 2.93 1.00 picture is aesthetical ly acceptable .846 .87 2.86 1.05 favorable first impression .820 .87 2.55 1.01 pictures gain attention .789 .87 2.52 1.06 picture fits with other pictures on the page .737 .88 3.21 .92 picture is technically acceptable .716 .88 3.32 .90 picture depiction is accurate .686 .88 3.16 1.00 picture fits into given area .645 .89 3.30 .99 graphics support text content .593 .90 2.52 1.05 Factor 2 Text Presentation 3.75 11.73 .75 3.28 .92 headings ID the subject matter .681 .76 3.03 .97 explicit text .660 .73 3.20 .92 well organized text .646 .67 3.29 .96 text in visually distinct segments .629 .67 3.47 .90 short blocks of text .527 .70 3.42 .87 Factor 3 Visual Text 2.08 6.48 .84 3.14 1.04 lines of text in 8 10 words .873 -3.06 1.06 column width appropriate .837 -3.21 1.01 Factor 4 Graphic/Text Support 1.29 4.04 .97 1.71 1.07 captions placed near graphics .938 -1.71 1.08 captions support graphics .933 -1.70 1.06 Factor 5 1.66 5.18 64 3.23 1.29 loss of text overview .842 .50 2.79 1.10 picture loss of overview .752 .51 3.02 1.10 white space appropriate use .435 .63 3.08 1.56 picture overload .430 .63 3.38 1.02 Factor 6 1.50 4.70 .55 2.92 .95 hotlinks imbedded in text .685 .44 2.10 1.20 same information in different formats .601 .48 2.20 1.11 typographic queuing .568 .47 2.23 1.00 color pictures preferable to black & white .467 .50 3.61 .89 Factor 7 1.27 3.96 .59 3.64 .77 realistic colors used i n graphics .632 -3.63 .81 consistent color scheme .590 -3.64 .72 Factor 8 1.06 3.31 .54 3.33 .88 multiple type styles .620 .55 3.30 .84 font size appropriate .568 .41 3.2 .82 leading / interlinear spacing .470 .33 3.46 .93 Extrac tion Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. Rotation converged in 14 iterations. Factor 1 = graphic presentation; Factor 2 = text presentation; Factor 3 = visual text; Factor 4 = graphics/text suppor t Cronbach alphas are in bold Factors 5 through 8 exhibited low reliability were not retained.

PAGE 96

86 contribution and elevated alphas resulting from inclusion of the item short blocks of text in factor 2, and its removal from factor 3, were sufficient to justify those actions. The alpha for the subscales graphic presentation text presentation visual text and graphic/text support was greater than .7, indicating that they were a reasonably good measure of the multi dimensional construct. Research Question 4 To answer Research Question 4 Do regional or geographic differences exist in the use of message design elements on CVB web sites? an analysis of variance by region was conducted Significant differences wer e found between world region s on the use of nine message design variable s: font size appropriate, picture loss of overview, pictures gain attention, favorable first impression, picture depiction is accurate, picture is aestheti cally acceptable, picture fits into given area, picture fits with other pictures on the page, and consistent color scheme A Bonferroni corrected post hoc test revealed that CVB web sites from Canada were significantly different from those in three other regions on six message design items. Oceania/South Pacific web sites also exhibited differences with those from three other regions on six ite ms. In these comparisons, Canadas means were lower than the other regions. Conversely, the means for Oceania/S outh Pacific were higher than the other regions Table 4 8 includes all the significant regional differences. The analysis of variance revealed a significant group difference on the factor graphic presentation F (1, 587) = 3.319, p = .001. No significant differences were found for the factors text presentation F (1, 587) = 1.236, p =

PAGE 97

87 .275; visual text F (1, 587) = .968, p = .460; or graphic/text support F (1, 587) = 1.499, p = .154. A s depicted in Table 4 9 Bon ferroni post hoc test indicated significant differences on the factor graphic presentation between Canada and six world regions (Africa, Central/South America, Europe, Caribbean, United States and Oceania/ South Pacific ) There were no significant differences between world regions for the message design factors graphic/text support text presentation or visual text Table 4 8 Message design elements exhibiting significant differences b etween r egions Message Design Element World Region Mean World Region Mean Sig 95% Confidence Interval consistent color scheme Asia 3.30 Oceania/South Paci fic 3.94 0.02 1.22 0.05 Africa 3.05 0.03 1.79 0.05 favorable first impression Canada 2.13 Oceania/South Pacific 3.03 0.00 1.62 0.18 pictures gain attention Canada 2.06 Oceania/South Pacific 3.00 0.00 1.70 0.18 font size appropriate Canada 2.91 United States 3.33 0.02 0.80 0.03 Africa 3.35 0.04 1.81 0.01 picture is aesthetically acceptable Canada 2.44 Oceania/South Pacific 3.23 0.03 1.53 0.04 Africa 3.59 0.04 1.74 0.02 picture depiction is accura te Canada 2.72 Oceania/South Pacific 3.48 0.02 1.47 0.05 picture fits into given area Canada 2.93 United States 3.41 0.04 0.95 0.01 picture loss of overview Europe 2.64 United States 3.25 0.00 1.01 0.22 consistent color scheme Middle East 3.15 Oceania/South Pacific 3.94 0.04 1.54 0.02

PAGE 98

88 Table 4 9 Graphic presentation sig nificant diff erence by region W orld R egion mean Sig. Canada (mean = 2.54) Africa 3.29 .00 Oceana/South Pacific 3.18 .0 2 Central/South America 3.08 .04 Caribbean 3.06 .0 5 Europe 2.96 .03 United States 2.94 .01 Asia 2.89 .49 Middle East 2.69 .99 The mean difference is significant at the .05 level. Other Variables Related to the Success of Destination Message Delivery The nature of several important web site features did not permit measu rement as scale items. Download time, readabili ty and legibility of web sites were analyzed separately, using software expressly designed to examine these web site elements Download time for each web site was measured and means were calculated for regions. Although the means of the nine world regions were not significantly different, African web sites delivered the fastest download time (Table 4 1 0 ) followed by Central/South America and the Caribbean. The six remaining regions all had download times in excess of o ne second. Readability was measured using the Flesch Kincaid Grade Level readability scale. The Middle East mean score was 5.79, indicating that between five and six years of schooling would be required to understand the textual content on Middle Easter n web sites. Mean scores for Canada, Africa, and Asia indicated that between six and seven years of education were necessary, while

PAGE 99

89 Table 4 1 0 Broadband download times by world region Region N Std. Deviation Std. Error Mean (in seconds) Canada 55 0.65 0.09 1.38 Middle East 13 1.02 0.28 1.33 Oceania/South Pacific 30 0.89 0.16 1.27 Asia 29 1.17 0.22 1.26 Europe 103 0.91 0.09 1.22 United States 270 0.82 0.05 1.18 Caribbean 30 0.55 0.10 0.95 Central/South America 24 0.70 0.14 0.94 Africa 18 0.63 0.15 0.87 mean 0.83 0.03 1.19 scores for the United States and all other regions signaled that over seven years of education would be required (Table 4 1 1 ). There were no significant differences between world regions for the legibility of the text and headings. Each world region achieved an acceptable mean (Table 4 1 2 ) score for legible text where scores closer to 1 indicate greater legibility than scores closer to 2. African web sites scored the highest for legible text text/backgroun d color difference and text brightness difference Table 4 1 1 Flesch Kincaid grade level readability scores by region. Region N SD Mean Central/South America 24 3.42 7.49 Caribbean 30 1.72 7.19 Oceania/South Pacific 31 1.63 7.17 Europe 105 2.16 7.09 United States 272 4.06 7.08 Asia 30 2.71 6.97 Africa 16 1.88 6.93 Canada 55 2.16 6.81 Middle East 13 2.80 5.79 mean 3.24 7.05

PAGE 100

90 Table 4 1 2 Legibility of text by region. Region N legible text text color difference* text brightness difference** Africa 17 1.06 652.18 221.00 Canada 55 1.13 639.42 211.82 Caribbean 29 1.14 644.52 220.62 Middle East 13 1.15 638.46 218.08 United States 276 1.15 651.74 219.88 Oceania/South Pacific 31 1.16 639.65 214.97 Asia 30 1.17 630.20 212.97 Central/South America 25 1.20 613.56 206.68 Europe 108 1.22 619.22 207.37 mean 1.16 640.54 215.66 *scores >500 are ac ceptable **scores >125 are acceptable The textual content was assessed for three types of color vision impairment. There were no significant differences found between world regions for color vision impairment, with each of the nine world regions achievin g acceptably high scores (Table 4 1 3 ). Table 4 1 3 Color vision impairment mean scores for text by region Region N text protanopia text deuteranopia text tritanopia Asia 30 1.17 1.17 1.13 Africa 17 1.06 1.12 1.06 Central/South America 25 1.20 1.16 1.16 Canada 55 1.18 1.18 1.20 Europe 108 1.24 1.24 1.24 Caribbean 29 1.14 1.21 1.10 Middle East 13 1.15 1.15 1.15 United States 277 1.16 1.17 1.17 Oceania/South Pacific 31 1.19 1.19 1.16 mean 1.18 1.18 1.17 The overall legibility of headi ngs appeared to be acceptable for each world regions, although four regions had unacceptably low scores (below 500) for

PAGE 101

91 heading color difference (Table 4 1 4 ). There were, however, significant differences found between text and heading color differences ov erall (Table 4 1 5 ), and text and heading brightness differences overall (Table 4 1 6 ). Table 4 1 4 Legibility of headings by region. Region N legible headings heading color difference* heading brightness difference** Asia 30 1.50 497.13 168.97 Africa 17 1.47 484.00 166.47 Central/South America 25 1.36 533.16 176.88 Canada 54 1.56 493.91 165.39 Europe 107 1.37 549.47 186.61 Caribbean 30 1.40 511.97 174.63 Middle East 13 1.46 485.46 159.62 United States 272 1.42 522.12 178.65 Oceania/South Pacifi c 31 1.48 504.84 173.13 mean 1.43 520.27 176.99 *scores >500 are acceptable (italics indicate unacceptable scores) **scores >125 are acceptable Table 4 1 5 Significant difference between text and heading color difference. Variable t df Sig. (2 tailed) Mean Difference 95% CI Lower 95% CI Upper text color difference 23.81 583 .000 140.54 128.94 152.13 heading color difference 2.94 574 .003 20.27 6.74 33.80 p is significant at .05; Test value = 500 Table 4 1 6 Significant difference bet ween text and heading brightness difference. Variable t df Sig. (2 tailed) Mean Difference 95% CI Lower 95% CI Upper text brightness difference 46.89 583 .000 90.66 86.9 94.5 heading brightness difference 21.50 573 .000 51.99 47.24 56.74 p is signif icant at .05; Test value = 125

PAGE 102

92 The headings were assessed for three types of color vision impairment. There were no significant differences found between world regions for color vision impairment, with each of the nine world regions achieving acceptably high scores (Table 4 1 7 ). Table 4 1 7 Color vision impairment mean scores for headings by region. Region N Headings Protanopia Headings Deuteranopia Headings Tritanopia Asia 30 1.53 1.53 1.47 Africa 17 1.59 1.59 1.47 Central/South America 25 1.44 1.40 1.40 Canada 54 1.52 1.59 1.67 Europe 107 1.40 1.42 1.41 Caribbean 30 1.47 1.50 1.50 Middle East 13 1.62 1.62 1.54 United States 272 1.45 1.48 1.46 Oceania/South Pacific 31 1.55 1.55 1.48 mean 1.46 1.49 1.47 Color vision impair ment for both text and headings was compared for regional and overall differences. There were no significant differences found between world regions for protanopia, deuteranopia or tritanopia. There were, however significant differences found for the CVB w eb sites overall with respect to differences in text vs. headings fo r protanopia (Table 4 18 ), deuteranopia (Table 4 1 9 ) and tritanopia (Table 4 2 0 ). Table 4 18 Text / headings difference: Protanopia. Variable T df Sig (2 tailed) Mean Difference 95% CI Lower 95% CI Upper Protanopia (text) 11.24 584 .00 .18 .15 .21 Protanopia (headings) 22.40 578 .0 0 .46 .42 .51 p is significant at .05; Test value = 1

PAGE 103

93 Table 4 19 Text / headings difference: Deuteranopia. Variable T df Sig (2 tailed) Mean D ifference 95% CI Lower 95% CI Upper Deuteranopia (text) 11.43 584 .00 .18 .15 .21 Deuteranopia (headings) 23.59 578 .0 0 .49 .45 .53 p is significant at .05; Test value = 1 Table 4 2 0 Text / headings difference: Tritanopia. Variable T df Sig (2 t ailed) Mean Difference 95% CI Lower 95% CI Upper Tritanopia (text) 11.11 584 .0 0 .17 .14 .21 Tritanopia (headings) 22.63 578 .0 0 .47 .43 .51 p is significant at .05; Test value = 1 Research Question 5 To answer Research Question 5 What web page ele ments do web design professionals believe to be most important for destination message delivery? a Delphi study using an international panel of web page design experts was conducted. After being assigned fictitious names, panelists were asked to ascribe i mportance ratings to a list of instructional message design elements, and then discuss their item rankings in an effort to reach consensus. Items were scored on a four point scale where 1=not very important, 2=somewhat important, 3= very impor tant, and 4=critically important. The final item importance rankings (Table 4 2 1 ) illuminated areas of agreement, and disagreement among the panel members. Of the 41 web page items under consideration, unanimous consensus was reached for eight items.

PAGE 104

94 Table 4 2 1 Delphi panel importance ratings of message design elements. Text / Graphics Elements Mr. Brown Mr. Red Mr. Blue Mr. Pink Mean Whitespace (appropriate use) 4 4 4 4 4 .00 Text color/background contrast & brightness 4 4 4 4 4 .00 Favorable first impression 4 4 4 4 4 .00 Legible font (san serif) 3 4 4 4 3.75 Font size appropriate 4 4 3 4 3.75 Graphics (charts) easy to read 3 4 4 4 3.75 Picture is aesthetically satisfactory 2 4 4 4 3.5 0 Picture fits into the given area 3 4 3 4 3.5 0 Consis tent color scheme 3 4 3 4 3.5 0 Procurement (download) time 2 4 4 4 3.5 0 Running text in lowercase letters 3 4 2 4 3.25 Visuals & graphics gain attention 2 4 3 4 3.25 Picture is technically acceptable 2 4 3 4 3.25 Meanings of icons immediately obvious 2 4 3 4 3.25 Information w/color also w/out color 3 3 3 4 3.25 Headings attract attention 3 3 3 3 3 .00 Familiar font 4 3 2 3 3 .00 Embedded hotlinks 2 4 3 3 3 .00 Use of readable language 3 2 3 4 3 .00 Well organized text 3 3 3 3 3 .00 Explicit text 3 3 3 3 3 .00 Graphics support and augment textual content 2 3 3 4 3 .00 Captions placed nearer, rather than farther, to corresponding graphics 3 3 2 4 3 .00 Headings identify the subject matter 2 3 2 4 2.75 Leading / Interlinear spacing (appropriate use) 3 3 2 3 2.75 Multiple type styles (appropriate use) 3 2 2 4 2.75 Captions support graphics 1 3 3 4 2.75 Picture depiction is accurate 3 4 4 2.75 Graphics used when messages are complex 3 3 3 2 2.75 Text in visually distinct segments 3 2 3 2 2.5 0 Shor t blocks of text 2 2 2 4 2.5 0 Minimum loss of overview (scrolling) 2 1 3 4 2.5 0

PAGE 105

95 Table 4 2 1 Continued. Text / Graphics Elements Brown Red Blue Pink Mean Avoid jargon, slang and other unfamiliar words 2 4 3 1 2.5 0 Color media preferable to black an d white 2 3 1 4 2.5 0 Minimum loss of overview (scrolling) 2 2 2 2 2 .00 Same information is delivered in different formats 3 2 2 1 2 .00 Picture fits in with the other pictures on the page 3 1 3 1 2 .00 Column width 2 2 2 1 1.75 Abbreviations, acronyms a nd initialisms (appropriate use) 2 1 3 1 1.75 Picture overload 2 2 3 1.75 Lines of text in 8 to 10 words 2 2 1 1 1.5 0 Provide metaphors or quotes to make the unfamiliar familiar 1 1 3 1 1.5 0 Content generalized to specific 2 2 1 1 1.5 0 Realistic co lors used in graphics 1 1 2 2 1.5 0 Highlighted key words 1 1 2 1 1.25 Macrosignals (underline, italics, bold) 1 1 1 1 1 .00 All of the panel members expressed a belief that whitespace, text color/background contrast and brightness and favorable firs t impression were critically important design elements. Panel members were also united in casting loss of overview as somewhat important, and macrosignals as not very important. Headings attract attention, well organized text, and explicit text were de emed very, although not critically, important. Near unanimous agreement of high importance was reached on seven additional items, legible font, font size appropriate, graphics/charts easy to read, picture is aesthetically satisfactory, pictures fits into the given area, consistent color scheme and procurement

PAGE 106

96 (download time). Of lesser importance were the variables running text in lowercase letters, visuals and graphics gain attention, picture is technically acceptable, meanings of icons immediately ob vious, Information w/color also w/out color, headings attract attention, familiar font, embedded hotlinks, use of readable language, well organized text, explicit text, graphics support and augment textual content, and captions placed nearer, rather than f arther, to corresponding graphics Considered to be somewhat important were the variables headings identify the subject matter, leading / interlinear spacing, multiple type styles, captions support graphics, picture, depiction is accurate, graphics used wh en messages are complex, text in visually distinct segments, short blocks of text, minimum loss of overview (scrolling), avoid jargon, slang and other unfamiliar words, and color media preferable to black and white Delphi panel members considered column w idth, abbreviations, acronyms and initialisms, picture overload, lines of text in 8 to 10 words, provide metaphors or quotes to make the unfamiliar familiar, content generalized to specific, realistic colors used in graphics, highlighted key words, and the aforementioned macrosignals to be of relatively minimal importance with respect to the other web page elements under discussion. Summary Data from 588 CVB web sites were analyzed to examine the extent of the use of instructional message design principles. While many message design items were nearly ubiquitous, others were little used. Three items appeared on less than 25% of the CVB sites; six appeared on less than 50%. All other

PAGE 107

97 message design items, however, appeared on over 75% of the web sites visited Factor analysis produced a four dimension model of web design factors that describe the construct of an instructional environment for destination message design. The four dimensions were graphic presentation text presentation graphic/text support and visual text There was a significant difference found between world regions for the factor graphic presentation Finally, experts in web design were called on to determine which variables they considered important in web design. In general, the Delphi panel considered the variables that dealt with the visual aspects of text and graphics to be the most important. Chapter 5 presents an interpretation of these results and implications for both practice and further researc h.

PAGE 108

98 CHAPTER 5 DISCUSSION In Chapter 4, the results of the data analysis were presented. Chapter 5 provides a di scussion of these analys e s, linkages to theory and findings from previous research, and implications for practice and future research. Discussion of the Results This study examined the use of instructional message design principles in the design of CVB we b sites around the world. Specifically, the nature and frequency of use was evaluated, and web sites were compared to detect regional differences. Web design experts were consulted to determine which elements were considered important to design professio nals. The following sections provide an in depth discussion of the results. Appearance of Message Design Elements For Research Question 1 Which elements of message design appear on CVB web sites? descriptive statistics revealed that, to a varying degree, each of the 588 CVB web sites contained some of the message design elements under review in this study (Table 4 5 ). The most pervasive message design practices were picture fits with other pictures, same information presented in different formats, conte nt written from generalized to specific, picture depiction is accurate, picture fits into given area, picture is aesthetically satisfactory, picture is technically acceptable, graphic overview (scrolling), graphics gain attention,

PAGE 109

99 text in visually distinct segments, short blocks of text, picture overload, lines of text in 8 to 10 words, realistic colors used in graphics, explicit text, color media preferable to B&W, broadband (download speed), running text lower case letters, column width, well organized te xt, leading / interlinear spacing, headings identify the subject matter, favorable first impression, legible headings, headings attract attention, legible font, consistent color scheme, familiar font font size appropriate, multiple type styles, loss of te xt overview, legible text, and whitespace (appropriate use) all appearing on more than 500 web sites. Many message design (30) items were nearly ubiquitous, appearing on more than 95% of the web sites assessed. Six of these, familiar font, font size appr opriate, multiple type styles, loss of text overview, legible text, and appropriate use of whitespace appeared on 99% of the CVB sites visited. All (100%) of CVB web sites use at least some of the instructional message design practices recommended to ease the cognitive load of web site visitors. This result is not surprising, as many message design elements had their genesis in print media and are familiar to both print and web authors alike, albeit not as methods to reduce cognitive load but as style guide s for best writing practices. Although writing for the web is different from writing for print media (e.g., scannable text, etc.) many of the same conventions developed for print have been adapted for electronic media. In each case, content is created fo r readers or viewers who have become accustomed to familiar presentation styles, whether in print or on the web. Most graphic designers follow traditional typographic rules for constructing text and displaying graphics. As a result, readers have become

PAGE 110

100 accustomed t o similar standards in typesetting, a familiarity that has become a reader expectation (Keller & Burkman, 1993). Participants in the Delphi study, with their emphasis on visual impact, reiterated the importance of many of these commonly used elements. Fo r example, several participants emphasized the necessity of a favorable first impression or as one participant expressed, I dont stay if I get a bad first impression. If they can t do that right, whats next? It just gets worse. Specific common el ements were considered essential, as well For example, whitespace is something everyone notices without knowing it. It registers quietly. Your avera ge user never notices. But it a ffects them anyway. Many of the other elements that were most commonly found on CVB sites were also rated as critically important or very important by the Delphi panelists. Some message design elements appeared on relatively few sites. The use of metaphors to make the unfamiliar familiar (appearing on only 41 o f 588 CVB sites ) information presented with color also presented without color (143 sites) and use of graphics when messages were complex (138 sites) were the least used me ssage design techniques. Interestingly, these three practices all share a redundancy component, as each of these three variables deals with providing a secondary mechanism for content delivery. Winne (1995) suggested that learning improves when the same information is delivered in different formats, but the results of this study indicate that the utilization of multiple formats for web content has been underused on CVB web sites.

PAGE 111

101 Members of the Delphi panel considered two of these little used elements as unimportant: metaphors to make the unfamiliar familiar and information presented with color also presented without color The third element, use of graphics when messages were complex which appeared on only 138 of the 588 sites, was considered very impo rtant by three of the four panel members and somewhat important by the fourth There are several possible explanations for the failure of many CVB site designers to include this important element. Sites with no complex messages would have no reason to include graphics for this purpose ; there was no distinction on the scale between sites that did not employ graphics when messages were complex and those that had no complex messages A potentially problematic explanation for the omission of graphics when messages were complex is that designers may not have considered their messages to be complex. This may occur because the web designer may be completely disconnected from the content of the site; that is, the web designer may be more concerned with making the site look good and making it functional than making it an effective conveyor of complex messages. This assumption is supported by the responses from the Delphi panel members. A critical function of CVB web sites is to convey information in a way th at will promote the users retention. According to cognitive load theor ists ( Sweller, 1994; Sweller, Van Merrinboer, & Paas, 1998), this can be most efficiently accomplished by reducing the demands on working memory. Content should be presented to focus the users working memory on relevant information and to avoid taxing it with irrelevant or distracting features The widespread use of

PAGE 112

102 numerous effective message design elements indicates that CVB web designers are at least somewhat aware of the need to support users retention of information. The responses of the Delphi panelists indicate that web professionals have accepted many of these elements as essential. However, even several of the elements that were considered unimportant by the Delphi panelist s were found on most of the web sites examined. For example, macrosignals (underline, italics, bold) were considered the least important by the panelists, yet 84.7% of the sites examined employed this element. Similarly, highlighted key words were used b y 80.1% of the CVB web sites. Effectiveness of Message Design Elements on CVB Web Sites Research Q uestion 2 Which message design elements do CVB web sites employ effectively to deliver destination messages ? was answered using the mean scores from the raters checklists Although, upon examination of the frequency data, i t became evident tha t the message design elements were commonly used, the quality of their use was quite variable The most striking aspect of the variability in quality was that the mean scores for visual impact related items tended to be high, but the mean scores for conte nt delivery related items tended to be low. CVB web sites use of text ( e.g., font size, familiarity and legibility, short blocks of text well organized text, distinct segments of text ) and graphics ( e.g., realistic colors, technical acceptability of pic tures, use of white space) was consistently rated very high (3.0 or higher). This indicates that CVB web sites

PAGE 113

103 are succeeding in the area of visual impact. In contrast, items related to making the destination message accessible and understandable tended to be rated lower. For example, items such as captions support graphics macrosignals (bold, italics, underline), imbedded hotlinks for key items and content generalized to specific were all rated below 2.25. While CVB sites tend to be visually appealin g, the poor ratings on these items indicate that they may be failing in their primary mission of conveying a destination message. This finding is consistent with the Delphi panelists emphasis of the imp ortance of visual impact The panelists rated most o f the content delivery related items as unimportant or only somewhat important. This disconnect between the focus of web designers and the mission of the CVB is problematic. Factor Analysis Research Q uestion 3 How many underlying instructional message design conditions that enable desired learning outcomes are present on CVB web sites? was answe red through the data reduction technique of factor analysis of the scale variables to illuminate the conditions or factors that explain the pattern of correlations within the CVB dataset. Factor scores were saved as new variables and used to int erpret what the factors represented conceptually. Four factors were retained in the final model. The first factor, graphic presentation contained the items p icture is aesthetically acceptable f avorable first impression g raphics gain attent ion picture fits with other pictures on the page picture is technically acceptable picture depiction is accurate picture fits into given area and graphics support the text All of the items comprising the

PAGE 114

104 factor gra phic presentation pertain to the visual display of graphic content. The alpha for this factor was quite high (alpha = .89 ), indicating that the above items were a reliable measure of the presentation of graphic content. This factor represents an i mportant aspect of message design, because recall of pictures is better than recall of text (Gagne & Rower, 1969) and visuals and graphic images are also perceived more rapidly than texts (Fleming & Levie 1978). This factor is also critical in presen ting a favorable first impression. Learners are more likely to feel confident and motivated about materials that convey a favorable first impression than those that do not. Members of the Delphi panel agreed on the importance of creating a favorable first impression, calling it a critical concern for designers. An often repeated opinion on the function of visual images concerns getting attention and maintaining attention (Pettersson, 1998). Like the items that comprise graphic presentation, most motivat ional graphic principles relate to gaining and maintaining attention (Keller & Burkman, 1993). Witt (1981) explained that, although pictures and graphics may not carry the main weight of a message, they are useful in attracting the direct attention of the learner. Attention is the mechanism that determines which stimuli will be engaged in and which stimuli will be ignored (Carlson, 1993). Mayer (1993) stated that for learning to occur, the learner must first pay attention to the relevant information. To a ccomplish this goal, instructional designers must effectively draw the learner's attention to the relevant information.

PAGE 115

105 The Delphi panelists were somewhat concerned with graphics, as they rated several elements related to graphics (e.g., graphics (charts) easy to read, picture is aesthetically satisfactory, picture fits into the given area ) as very important or critically important. Interestingly, the Delphi panelists were less concerned with whether the graphics contributed substantively to the messa ge of the site They rated several items related to the content of the picture (e.g., graphics used when messages are complex, captions support graphics, picture fits in with the other pictures on the page ) as less important. The focus of these web desig n experts was very clearly on the visual impact and functionality of graphics, not on the message delivery aspects For example, one panelist explained, People dont read most captions. Unless I need a caption to identify someone, I avoid using them. As a rule. Again, the disconnect between web designers focus on visual impact and functionality of a site and the CVBs mission of conveying a destination message may be at odds. The appropriate use of textual features can increase attention and influence the motivation to learn, (Keller & Burkman, 1993). The second factor, text presentation was comp o sed of five message design items that pertained to the visual display of text. These factors were headings identify the subject matter, explicit tex t, well organized text, text in visually distinct segments, and short blocks of text This factor was important because the formatting of text based content can determine whether it is perceived as easy or difficult t o use (Keller & Burkman ). Relatively short blocks of text and concise messages tend to suggest fairly easy reading, and that initial easy to read perception builds confidence and

PAGE 116

106 maintains attention (Keller & Burkman ). Designers can also create the perception of easy reading by dividing lengthy blocks of t ext into visually distinct segments (Keller & Burkman ; Smith & Orr, 1985; White 1987). The third factor, visual text (alpha = .7 7) pertains to the layout of text on the page and was comp o sed of three items lines of tex t in 8 10 words column width appropriate and short blocks of text Research indicates that lines of text should be composed of eight to ten words in 12 point font to make text easier to read, maintain learner attention and increase confidence, and that lines of text should also be presented in a fairly open, rather than constrained display (Keller & Burkman, 1993). The Delphi panelists rated many of the items related to text presentation and visual text as critically important or very important. In particular, items related to the font size and legibility and the contrast between text and background colors were considered critical. Again, these items relate to the visual impact of the web site, not to its content. In contrast, items such as organizing content from generalized to specific and using macrosignals (bold, italics, underline) or highlighting of key words to emphasize important concepts were considered not important at all. The panelist s considered text elements such as te xt in visually distinct segments and short blocks of text only somewhat important. These elements contribute significantly to the readability of text and, therefore, to destination message delivery. The final factor, graphic/ t ext s upport was composed of two items, each pertaining to the use of captions ( captions support graphics captions placed near

PAGE 117

107 to corresponding graphics ) within web page design. The high alpha (.9 7) for this factor is not surprising, as a high correlation betw een the two caption items should be expected. This finding is helpful for web message designers, as physical integration of visual and verbal content can decrease cognitive load and increase learning (Chandler & Sweller, 1991). Sound design ensures that learning is enhanced when text is placed nearer, rather than farther, to corresponding pictures and graphics (Mayer, 1993). When both text and graphics convey exactly the same information, users may become bored or distracted. Text and graphics should be c omplementary, offering slightly different views of related information to promote learning (McFarland, 1995). As previously mentioned, the Delphi panelists were not particularly concerned with the use of captions. In fact, one member of the panel avoids using them altogether. Regional Differences in the Use of Message Design Ele ments Research Question 4 Do regional or geographic differences exist in the use of message design elements on CVB web sites? was answered by an analysis of variance of the nine world regions contained in the CVB data set. Given the uneven diffusion of th e Internet, regional differences were expected. Tufte (1990) noted that the principles of information design are universal and are not aligned with any particular language or culture. Still, there was a significant difference found between world regions on the use of nine message design variables: font size appropriate, picture loss of overview, pictures gain attention, favorable first impression, picture depiction is accurate, picture is aesthetically acceptable, picture fits into given area, picture fits with other

PAGE 118

108 pictures on the page, and consistent color scheme. The differences were related to lower scores of sites in the Canada region and higher scores of sites in the Oceania/South Pacific Region. These differences may be due to the differences in the degree of reliance on touri sm in the respective regions. That is, the Oceania/South Pacific region relies more on tourism which dominates the economies of many countries in the region. The remote locations of many of the island nations create more reliance on the Internet to commu nicate with potential visitors than some of the other regions. In comparison to islands in Oceania/South Pacific tourism is a less integral part of the Canadian economy. Also, with the US as a primary source of tourists, Canadian CVBs have other outlets available for communicating their destination messages. A further analysis of variance on the four factor subscales applied to the nine world regions revealed a significant difference on one of the factors, graphic presentation This differ ence was between Canada and other world regions. Given the lesser importance of tourism to the economy in Canada compared with some other world regions, it would logically follow the investment in CVBs and, in particular, CVB web sites in Canada would be less. Graphic presentation is the factor most dependent on professional designers. The other factors, which are more closely aligned with well established print conventions, consist of elements that can be generated by designers with less expertise. Hig h quality graphic design requires more skill. Therefore, one explanation for the differences on this factor could be differences in CVB budget allocations to web design in Canada. Additional study would be required to confirm or refute this

PAGE 119

109 explanation. T here were no significant difference s found between world regions for the remaining three subscales Other Variables Related to the Success of Destination Message Delivery The web site visitors exposure to site content can be influenced by a plethora of technical issues, such as the speed of the ISP, computer settings and capabilities, and web browser preferences and settings. The combination of these features defines the technical environment that affects the display of information available to site visitors (Hodkinson & Kiel, 2003). Several important web site features were not measured as scale items. Download time, readability, and legibility of web sites were measure d using specialized software useful for examining such elements. Scores for these items were compared to detect any regional differences. Web user impatience with overlong waiting times is a problem that can, in part, be solved by good design. Download ti me is a critical area of concern for web users, and extended waiting times are not suffered gladly in the wired environment. Research has indicated that tolerable waiting time for web users is about two seconds, after which users begin to search for alte rnate sites (Fui Hoon Nah, 2004). As broadband download time for the CVB web sites in this study averaged 1.19 seconds, there does not appear to be a widespread problem in this area. However, 13.9% of the web sites measured reported download times in exces s of two seconds, and a few sites ranged a high as 4.8 seconds. It should be noted that download time is not wholly a function of web site design and connection speed, and that download times for CVB web sites

PAGE 120

110 were recorded as a snapshot of conditions o n a particular day. For download time, there were no significant differences found between the regions. This was not unexpected, as the structure of the World Wide Web spans the regions, that is, each region is not an autonomous web with self contained sys tems comparable to other world regions. Significant differences between web site download time could be attributable to many factors, such as server speed or total net traffic. As such, world region may not be a valid parameter with which to assess downloa d speeds. Using readable language is an important message design consideration, as readable language helps gain and maintain learner attention (Keller & Burkman, 1993). Although readability typically refers to the ease with which the language in text is re ad and understood (Chall, Bissex, Conrad, & Harris Sharples, 1996), other aspects of text also contribute to its readability. Readability was measured using the Flesch Kincaid Grade Level readability scale. Hull (2004) reports that clear and simple languag e is recommended when designing for readers with different reading levels. Keller and Burkman (1993) recommended that designers consider their target learner's reading skills and interest in reading when designing text. There were no significant difference s between world regions in the readability of the web sites visited. The Middle East mean score was 5.79, indicating that nearly six years of schooling would be required to understand the textual content on Middle Eastern web sites. Mean scores for Canada Africa, and Asia were slightly higher (between six and seven years of education), while

PAGE 121

111 scores for the United States, Europe, Oceana/South Pacific, the Caribbean, and central/South America were all somewhat over seven years. Hull (2004) noted that th e average resident of the United States had an eighth grade reading level, therefore, a Flesch Kincaid score somewhat over seven would be appropriate for the American target audience. It should be noted that the Flesch Kincaid Grade Level readability scal e scores all rendered content, including navigational items, which could skew the results for a particular web site containing little text but a great deal of navigation. The legibility of the text and headings on CVB web sites was measured as a function of the foreground/background color contrast. There were no significant differences between world regions for legibility of the text or headings. Although the overall legibility of headings appeared to be acceptable for each of the world regions, four regio ns had unacceptably low scores (below 500) for heading color difference (Table 4 1 4 ). There were, however significant differences found between text and heading color differences overall (Table 4 1 5 ), and text and heading brightness differences overall (Table 4 1 6 ). Additionally, the data revealed that just over 16% of the CVB sites visited did not exhibit sufficiently legible text, and over 42% did not exhibit sufficiently legible headings. CVB sites with such poor legibility scores can hardly be said to be focused on Keller and Burkmans (1993) recommendations for maintaining learner attention or increasing confidence. In particular, design of text and headings that fail to meet legibility standards seems counterproductive in a message design context, and may reflect an approach to design that seeks to

PAGE 122

112 engage the eye at the expense of engaging the brain. During the Delphi study, one of the panel members indicated that they would opt for less distinct text if it blended better with the overall page desi gn. Viewed from that perspective, it is not surprising that text, and headings in particular, could be produced with less emphasis on legibility than was optimal. The text and heading foreground/background color contrast was also assessed for three types of color vision impairment. There were no significant differences found between world regions for text color vision impairment, with each of the nine world regions achieving ac ceptably high scores (Table 4 14 ). Overall, however, the presentation of text and headings for the visually impaired varied greatly. For text, 17.7% of the web sites visited failed to provide adequate color conditions for site visitors with Protanopia, while 18.2% failed for Deuteranopia, and 17.3% failed for Tritanopia. For headi ngs, the results were much worse. For the CVB web sites visited, 45.7% failed to provide adequate color conditions for visitors with Protanopia, 48.3 failed for Deuteranopia, and 46.3 failed for Tritanopia. Hartley (1987) observed that Deuteranopia (red g reen) color blindness is the most common, and is more widespread among men than among women. Unfortunately, red and green are also used very often as key colors in symbols and graphics (Pettersson, 1998). The high level of failure to provide adequate col or conditions on CVB web sites is a message design problem that renders the site content inaccessible for many users.

PAGE 123

113 Importance of Message Design Elements to Web Design Professionals To answer Research Question 5 What web page elements do web design p rofessionals believe to be most important for destination message delivery? an online Delphi study was conducted using an international panel of web site design experts. Given the ever changing nature of the Internet and, in particular, of web design prac tices, consultation with practicing experts in the field was necessary to allow for more meaningful interpretation of the data and a better understanding of the implications of this study for professional practice. In addition to the relevant Delphi panel comments previously cited, a summary of the entire Delphi study is provided in the following sections. All of the Delphi panel members agreed that whitespace, text color/background contrast & brightness and favorable first impression were critically imp ortant design elements. Bevan (1998) cautioned that unnecessary whitespace may make scanning of the text more difficult, resulting in an imposition on gaining attention. Delphi panel members, however, viewed whitespace as part of creating a favorable fi rst impression and was important as the background tying everything together. Panelists, however, thought that creating a favorable first impression was more important, by far Although united in their feeling that text color/background contrast & br ightness was critically important, panelists were less certain of its absolute necessity, stating that its a tonal decision and that people will go with whatever looks better when opting for less distinct text if that is what fits better with overal l design.

PAGE 124

114 Items that rated very highly, although not unanimously critical, were legible font, font size appropriate, graphics easy to read, picture is aesthetically satisfactory, picture fits into given area, and consistent color scheme Panelists believed that a web site should have a professional appearance, echoing Kormiers (1999) statement that a beautifully designed, current, and informative web site can establish a companys presence and enhance the identity of the company. Panelists thought that having a consistent color scheme was a key factor. The panel reiterated that creating a favorable first impression was the most critically important item, and that all of the visual elements were tied into that objective. Keller and Burkman (1993) agreed, underlining the importance of designing text that creates a favorable first impression, as a favorable first impression enables the reader to scan the content and feel confident that they can easily read and understand the message. A favorable first impr ession may help to gain and keep attention and build confidence that readers can successfully navigate and use the content. Gaining and keeping the attention of the reader is also a useful method of reducing the imposition of nonlearning events on working memory. The next group of web page elements (in order of importance) includes: running text in lowercase letters, headings attract attention, familiar font, visuals & graphics gain attention, picture is technically acceptable, procurement (download) time, meanings of icons immediately obvious, imbedded hotlinks, use of readable language, well organized text, explicit text, minimum loss of overview

PAGE 125

115 (scrolling), graphics support and augment textual content, and captions placed nearer rather than farther to co rresponding graphics. One panel member made the distinction that scrolling up/down was acceptable, while scrolling from side to side was not, claiming that scrolling side to side is awful & should never happen. Another panel member cautioned that impact navigation and content are the critical design items, and that were splitting hairs on a lot of these, reflecting the statements of several researchers : ( a ) Jung and Butlers (2000) important characteristics of a successful web sites that included appearance, ease of navigation, design, and content ; ( b ) Bevans (1998) statement that successful web site development was contingent upon the developers awareness of graphic design and web usability ; and ( c ) that m ost web site developers do not consi der aspects that are specific to education and the teaching learning process (Montilva, Sandia, & Barrios 2002). Other graphic related items such as visuals and graphics gain attention, graphics support and augment textual content, and captions placed ne arer, rather than farther, to corresponding graphics were all deemed to be important for gaining attention. Interestingly, one panel member added that much of his opinion/ratings were subjective, and I dont guarantee that my answers would be the same t omorrow. He cautioned that design decisions had to be made in context, and warned that all my scores are as a rule but they all could change in context. Panel members were also united in casting loss of overview as somewhat important, and macrosignal s as not very important. Lastly, a panel member

PAGE 126

116 noted that If we screwed up any of these items, very few would be catastrophic. Whitespace was unanimously important, but if I didnt quite get that right, few would notice. In general, Delphi panel membe rs considered design for the web to be an almost wholly visual experience (Its on a monitor!). This viewpoint reflects the findings of Rayman Bacchus and Molina (2001), who noted that the success of leading tourism sites suggested a competitive advantag e for sites that were visually appealing and information rich and Pettersson (1999), who considered that message design principles could be best utilized as guidelines for the display of a visual messages. The experts believed that functionality and aesth etics were the most essential web site features, and admitted that they seldom thought about content delivery in the terms presented in this study. This finding is particularly important, because the mission of CVBs is to deliver destination messages ( Gal larza et al. 2002 ) If content on a CVB web site is not acquired or retained by the user because the web site designer is more concerned with aesthetic elements than content, the web site will not perform its function in helping the CVB achieve its missi on. Limitations and Delimitations of the Study Although the investigator rigorously adhered to predetermined evaluation criteria, inherent subjectivity by multiple raters in application of the measurement instrument was unavoidable. Although interrater re liability was established within acceptable limits for this study, a small degree of error variance between raters was present in the final results.

PAGE 127

117 The stratification of the sample into world regions was based on the source site (tourismbureaus.com). Th e evolving nature of the Internet subsequently affected the availabil ity of some sites in the sample, producing some experimental mortality or differential loss of cases due to web sites dropping out of the study on a potentially non random basis. Becaus e this study examined only CVB web sites, results may not be generalized to commercial tourism sites. Generalization of the findings to other types of informational sites may be similarly inappropriate. The study of instructional message design was devel oped, primarily, in western countries. Application of the tenets of message design to web sites developed for non western countries may present an imposition of cultural bias However, given that the study was conducted on only English language CVB sites the presumption that the target audience is predominantly western may be valid In a related vein, regional differences cannot be considered cultural differences because several of the regions included in the study are inhabited by people from many cu ltures. Web design is conducted without respect to national borders; that is, a web site representing a CVB may be created by a web designer in a different part of the world. Finally, although the data have been classified by region, regional classificati on does not imply regional causality. Implications for Practice Ultimately, the design of web sites for Convention and Visitors Bureaus (CVBs) must consider the role of the CVB and the goals of the users. CVBs help

PAGE 128

118 to generate tourism by providing informa tion in the form of a clear and appealing destination message. This study revealed that CVB web sites vary widely in their use of effective message design principles. Although numerous design elements appear to be ubiquitous (e.g., headings, use of a con sistent color schemes), others were rarely found (e.g., use of metaphors). This study yielded several important findings for CVB marketing personnel and web site designers. A wide range of expertise exists among CVB web designers and an apparent bias in favor of visual over content related elements exists among experts One the tenets of any type of design is that form follows function. If the function of a CVB web site is to deliver a destination message, that function cannot be achieved when message design precepts related to content are ignored. An understanding of the essential features of instructional message design would increase the likelihood that designers would employ these features. Many sites exhibited features such as prominent but irr elevant animations or other visual distractions, demonstrating a lack of understanding of the basic elements of effective message design. Getting attention and maintaining attention are two separate tasks. Maintaining attention is made more difficult by the use of distracting elements such as advertising, pop up window s, or music on web sites all of which increase cognitive load As cognitive load theorists suggest, maintaining attention is essential to promote retention of information.

PAGE 129

119 Several recomm endations for improving the design of CVB web sites emerged. Web sites should support the user by reducing cognitive load. This can be accomplished by making sites that are visually appealing, logically organized, and considerate of the user by incorpora ting the principles of instructional message design. Web designers should consider the elements of effective message design as more important than the technical aspects of the site. Although functionality is essential, the slick features many designers u se in an apparent attempt to look state of the art can actually serve as distractions to the user. Designers must remain focused on the purpose of the CVB Finally, CVBs should ensure that their destination message is conveyed in a way that web site user s will receive and retain the intended message. As both the Internet communications and tourism industries grow, it is likely that the design of CVB web sites will become an increasingly specialized field. Specialists in this area should recognize the f undamental difference between the purpose of CVB web sites and other sites. The purpose of a CVB site is to increase tourism to an area by educating potential visitors about the positive aspects of the destination. To accomplish this goal, implementation of effective instructional message design principles is essential. Implications for Future Research Although this study answered important questions about the use of message design principles on CVB web sites, it produced many questions, as well. Some of these questions resulted from the factor analysis and others were raised during the Delphi study.

PAGE 130

120 A logical follow up to the factor analysis would be to conduct a cluster analysis to determine how individual sites clustered on the four factors. A cluster analysis could enhance understanding of the issues contributing to the use of effective message design. Regional or population variables may be important within the clusters. Is the effective use of message design principles related to characteristics of the CVB? A study to examine this issue would consider factors such as the size of the population represented by the CVB; that is, are web sites representing countries more effective than those representing states, provinces, or municipalities? In additio n, is the effectiveness of the web site related to the CVBs budget? Is the popularity of the destination related to the effectiveness of the web site? Presenting an effective destination message is more critical for destinations that rely more heavily o n tourism for their economic vitality. To control for the wide variability in the size (i.e., number of pages and features) of web sites, this study was limited to examination of the lead page of the CVB sites in the sample. A follow up study to examine the functionality of sites beyond the lead page is warranted. That is, further research is necessary to determine the efficiency and ease of use of the navigation features, and the ease with which the user can find desired information. The experts in the Delphi study considered the functionality of sites to be of critical importance. Because navigation was outside the scope of this study, a follow up study on usability of CVB web sites may be warranted. Similarly, additional research on the accessibility of the sites would be useful.

PAGE 131

121 The experts in the Delphi study were also concerned with aesthetics of web sites. The items on the scale dealing with aesthetics were limited to variables that made the site more readable and made features of the site more easily distinguishable. The scale did not include items focused on how pleasing a site was to the eye or how appealing the text or graphics were to the viewer. For this reason, a follow up study examining the subjective preferences o f the user may prov e to be illuminating. Future research should examine the use of features that may distract the user. Many sites use animation or other eye catching elements to attract attention to particular features on the site. Many also include commercial advertisin g, which is likely used to offset the costs of creating and hosting the web site. A study of the relative effects of attractions versus distractions on CVB sites could provide further insight into what makes a site effective. Perhaps the most important di rection for future research is to include travelers in the rating of CVB web sites. A study that examined users capacity to retain information from sites exhibiting the various features included in this study would be particularly valuable.

PAGE 132

122 APPENDIX A INSTRUCTIONAL MESSAGE DESIGN ELEMENTS ASSESSED ON CVB WEB SITES Text Elements Item Reference Loss of overview (scrolling) Weitl, S, Kammerl, & Freitag, 2002, Dyson (2004). Whitespace (appropriat e use) Bevan, 1998, MacGregor & Lou (2004), Shneiderman & Plaisant, 2005. Text in visually distinct segments Keller & Burkman, 1993; Smith & Orr, 1985; White 1987; Allen & Eckols (1997), MacGregor & Lou (2004). Short blocks of text Keller & Burkman, 1993 ; Allen & Eckols (1997); McGovern, Norton & ODowd, 2002, Weitl, S, Kammerl & Freitag, 2002., Lee & Boling, 1999, MacGregor & Lou (2004). Running text in lowercase letters Henny, 1981; Poulton & Brown, 1968, Lee & Boling, 1999. Lines of text in 8 to 10 words Keller & Burkman, 1993, Lee & Boling, 1999, Dyson (2004), Shneiderman & Plaisant, 2005. Macrosignals (underline, italics, bold) Cisotto and Boscolo (1995), Lee & Boling, 1999. Text color/background contrast & brightness Pettersson (1991), Lee & B oling, 1999, Woodland & Szul (1999), MacGregor & Lou (2004), Hall & Hanna (2004), Shneiderman & Plaisant, 2005. Typographic queuing Pettersson (1991); Allen & Eckols (1997); McGovern, Norton & ODowd, 2002, Lee & Boling, 1999, MacGregor & Lou (2004). Hea dings attract attention Bevan, 1998, Cisotto and Boscolo (1995); Allen & Eckols (1997); McGovern, Norton & ODowd, 2002. Lee & Boling, 1999, MacGregor & Lou (2004).

PAGE 133

123 Headings identify the subject matter McGovern, Norton & ODowd, 2002, MacGregor & Lou (200 4). Highlighted key words Bevan, 1998; Allen & Eckols (1997), Wu & Yuan (2003). Leading / Interlinear spacing (appropriate use) Keller & Burkman, 1993, Dyson (2004), Shneiderman & Plaisant, 2005. Column width Keller & Burkman, 1993; McGovern, Norton & O Dowd, 2002, Dyson (2004). Abbreviations, acronyms and initialisms (appropriate use) Allen & Eckols (1997) Avoid jargon, slang and other unfamiliar words Allen & Eckols (1997); McGovern, Norton & ODowd, 2002. Provide metaphors or quotes to make the unf amiliar familiar Allen & Eckols (1997) Same information is delivered in different formats Winne (1995); Allen & Eckols (1997), Naidu & Jarvela (2005). Use of readable language Pettersson, 1999, Keller & Burkman, 1993; Allen & Eckols (1997); McGovern, No rton & ODowd, 2002, MacGregor & Lou (2004). Well organized text Keller & Burkman, 1993, Lee & Boling, 1999. Explicit text Keller & Burkman, 1993, MacGregor & Lou (2004). Content generalized to specific Weitl, S, Kammerl, & Freitag, 2002 Familiar fon t Keller & Burkman, 1993; McGovern, Norton & ODowd, 2002., Lee & Boling, 1999, Shneiderman & Plaisant, 2005. Legible font (san serif) Keller & Burkman, 1993; McGovern, Norton & ODowd, 2002., Lee & Boling, 1999, Shneiderman & Plaisant, 2005. Font size a ppropriate Keller & Burkman, 1993, Shneiderman & Plaisant, 2005.. Multiple type styles (appropriate use) Pettersson, 1999, Lee & Boling, 1999, Shneiderman & Plaisant, 2005. Captions support graphics Pettersson, 1985; McFarland, 1995, Lee & Boling, 1999.

PAGE 134

124 Graphic Elements Item Reference Loss of overview (scrolling) Weitl, S, Kammerl, & Freitag, 2002, Dyson (2004), Shneiderman & Plaisant, 2005. Visuals & graphics gain attention Fleming & Lev ie 1993, Lee & Boling, 1999. Favorable first impression K eller and Burkman, 1993, Shneiderman & Plaisant, 2005. Picture depiction is accurate Pettersson 1998 Picture is aesthetically satisfactory Pettersson 1998; McGovern, Norton & ODowd, 2002. Picture is technically acceptable Pettersson 1998 Picture fits into the given area Pettersson 1989 Picture fits in with the other pictures on the page Pettersson 1989; McGovern, Norton & ODowd, 2002. Picture overload Evans, Watson, and Willows,1987; Pettersson, 1998 Procurement (download) time Pettersson 1998 Gra phics used when messages are complex March, 1983, McGovern, Norton & ODowd, 2002. Graphics (charts) easy to read Pettersson, 1999, Keller & Burkman, 1993, MacGregor & Lou (2004). Meanings of icons immediately obvious Pettersson, 1998 Graphics support a nd augment textual content McFarland, 1995; Allen & Eckols (1997); McGovern, Winne (1995); Norton & ODowd, 2002. Captions placed nearer, rather than farther, to corresponding graphics Mayer, 1993, Robinson, Corlis, Bush, Bera & Tomberlin (2003). Consis tent color scheme McFarland, 1995, Lee & Boling, 1999. Information w/color also w/out color Pettersson, 1998

PAGE 135

125 Realistic colors used in graphics Keller & Burkman, 1993; Allen & Eckols (1997) Color media preferable to black and white Simonson 1984; Allen & Eckols (1997)

PAGE 136

126 APPENDIX B DELPHI PANEL WEB PAGE ELEMENTS IMPORTANCE CHECKLIST Not Important Somewhat Important Very Important Critically Important Dont Know Text Elements 1 2 4 5 1 Whitespace (appropriat e use) 1 2 4 5 2 Text in visually distinct segments 1 2 4 5 3 Short blocks of text 1 2 4 5 4 Running text in lowercase letters 1 2 4 5 5 Lines of text in 8 to 10 words 1 2 4 5 6 Macrosignals (underlin e, italics, bold) 1 2 4 5 7 Text color/background contrast & brightness 1 2 4 5 8 Headings attract attention 1 2 4 5 9 Headings identify the subject matter 1 2 4 5 10 Highlighted key words 1 2 4 5 11 E mbedded hotlinks 1 2 4 5 12 Leading / Interlinear spacing (appropriate use) 1 2 4 5 13 Loss of overview (scrolling) 1 2 4 5 14 Column width 1 2 4 5 15 Abbreviations, acronyms and initialisms (appropriate use) 1 2 4 5 16 Avoid jargon, slang and other unfamiliar words 1 2 4 5

PAGE 137

127 Not Important Somewhat Important Very Important Critically Important Dont Know Graphic Elements 17 Provide metaphors or quotes to make the unfamiliar familiar 1 2 4 5 18 Same information is delivered in different formats 1 2 4 5 19 Use of readable language 1 2 4 5 20 Well organized text 1 2 4 5 21 Explicit text 1 2 4 5 22 Content generalized to specific 1 2 4 5 23 Familiar font 1 2 4 5 24 Legible font (san serif) 1 2 4 5 25 Font size appropriate 1 2 4 5 26 Multiple type styles (appropriate use) 1 2 4 5 27 Captions support graphics 1 2 4 5 28 Loss of overview (scrolling) 1 2 4 5 29 Visuals & graphics gain attention 1 2 4 5 30 Favorable first impression 1 2 4 5 31 Picture depiction is accurate 1 2 4 5 32 Picture is aesthetically satisfactory 1 2 4 5 33 Picture is technically acceptable 1 2 4 5 34 Picture fits into the given area 1 2 4 5 35 Picture fits in with the other pictures on the page 1 2 4 5 36 Picture overload 1 2 4 5 37 Procurement (download) time 1 2 4 5 38 Graphics used when messages are complex 1 2 4 5

PAGE 138

128 Not Important Somewhat Important Very Important Critica lly Important Dont Know Graphic Elements 39 Graphics (charts) easy to read 1 2 4 5 40 Meanings of icons immediately obvious 1 2 4 5 41 Graphics support and augment textual content 1 2 4 5 42 Captions placed nearer, ra ther than farther, to corresponding graphics 1 2 4 5 43 Consistent color scheme 1 2 4 5 44 Information w/color also w/out color 1 2 4 5 45 Realistic colors used in graphics 1 2 4 5 46 Color media preferable to bl ack and white 1 2 4 5

PAGE 139

129 Explanation of Checklist Items Text Elements Loss of overview: main message is presented before the fold (w/ out scrolling) Whitespace: used to set page elements apart Text in visually distinct segment s: sections or paragraphs Short blocks of text: text is in 4 6 line blocks Running text in lowercase letters: no all caps Lines of text in 8 to 10 words Macrosignals: use of underline, italics, bold for emphasis Text color/background contrast & b rightness Typographic queuing: Headings attract attention: headings are larger, bolder, or more colorful than accompanying text Headings identify the subject matter Highlighted key words: highlighting for emphasis Leading / Interlinear spacing: appro priate space between lines Column width Abbreviations, acronyms and initialisms: meanings should be identified Avoid jargon, slang and other unfamiliar words Provide metaphors or quotes to make the unfamiliar familiar Same information is delivered in different formats: for example, directions for driving by car are accompanied by a map Use of readable language: appropriate grade level Well organized text : in sections, paragraphs, w/ headings, etc. Explicit text: message stated plainly and simply C ontent generalized to specific Familiar font Legible font: san serif font is used Font size appropriate: 12 pt font for running text Multiple type styles: appropriate use (no more than 3 per page) Captions support graphics Graphic Elements Loss of overview: can view without scrolling Visuals & graphics gain attention Favorable first impression Picture depiction is accurate Picture is aesthetically satisfactory Picture is technically acceptable: not blurry, etc. Picture fits into the given are a (without overlap) Picture fits in with the other pictures on the page

PAGE 140

130 Picture overload: too many pictures Procurement (download) time Graphics used when messages are complex Graphics (charts) easy to read Meanings of icons immediately obvious Grap hics support and augment textual content Captions placed nearer, rather than farther, to corresponding graphics Consistent color scheme Information w/color also w/out color: info that relies on color also explained in text Realistic colors used in grap hics Color media preferable to black and white: Hindenburg black and white picture is OK; Exxon Valdez color picture is preferable

PAGE 141

131 REFERENCES Allen, B. S., & Eckols, S. L. (Eds). (1997). Handbook of usability principles Center for Learning, Instruction, & Performance Technologies, San D iego State University, San Diego, CA. Alley, L. R. (1996). An instructional epiphany. Information Technology, 28 ( 2), 49 54. American Vocational Association, Inc. (2000). Now Showing: The Virtual College Tour. Techniques, 75 (1), 13. Assael, H. (1984). Co nsumer behavior and marketing action Boston: Kent. Babbie, E. (2001). The practice of social research (9 th ed.). Belmont, CA: Wadsworth Baines, A. (1998). Technology and tourism. Work Study, 47 (5), 160 163. Balakrishnan, R. (1991). Information acqui sition and resource allocation decisions. The Accounting Review, 66 (1) 120 139. Baloglou, S., & McCleary, K. W. (1999). A model of destination image formation. Annals of Tourism Research 26, 868 897. Barrett, P. (2001, March). Assessing the reliability of rating data revised Retrieved July, 2006, from http://www.pbmetrix.com/ Benson, P. J. (1985). Writing visually: Design considerations in technical publications. Technical Communications Journal, 32 207 216. Berge, Z. L. (2000). Components of the on line classroom. In R.E. Weiss, D.S. Knowlton & B.W. Speck (Eds.) Principles of Effective Teaching in the Online Classroom 23 28. Bevan, N. (1998). Usability issues in web site design. Retrieved October 2, 2004, from: http://www.npl.co.uk/npl/sections/u s Bieger, T. & Laesser, C. (2004). Information sources for travel decisions: Toward a source process model. Journal of Travel Research, 42 357 371.

PAGE 142

132 Business Wire. (2002, January 22). Measure the impact of web site improvements with WebCriterias enhan ced site analyst. Retrieved February 1, 2002, from http://www.businesswire.com/ Carlson, N. R., (1993). Psychology: The science of behavior (4 th ed.). Needham Heights, MA: Allyn & Bacon. Chall, J. S., Bissex, G L., Conrad, S. S., & Harris Sharples, S. (1996). Qualitative assessment of text difficulty: A practical guide for teachers and writers Cambridge, MA: Brookline. Chandler, P. & Sweller, J. (1991). Cognitive load theory and the format of instruction. C ognition and Instruction, 8 293 332. Cho, Y., & Fesenmaier, D. R. (2000). A conceptual framework for evaluating effects of a virtual tour. In D. R. Fesenmaier, S. Klein, & D. Buhalis (Eds.), Information and Communication Technologies in Tourism 2000 P roceedings of the International Conference in Barcelona, Spain, 314 323. Chou, C. (2002). Developing the e Delphi system: A web based forecasting tool for educational research. British Journal of Educational Technology, 33 (2), 233 236. CIAC (1998). US De partment of Energy Computer Incident Advisory Capability Information Bulletin I 034: Internet Cookies March 12, 1998. Retrieved May 30, 2006, from http://www.ciac.org/ciac/bulletins/i 034.shtml. Cisotto, L., & Boscolo, P. (1995, August). Effects of parag raph subtitles and underlining on written summaries Paper presented at the 6 th European Conference for Research on Learning and Instruction, Nijmegen. Cooper, G. (1998). Research into cognitive load theory and instructional design at UNSW School of Educa tion Studies, the University of New South Wales, Sidney, NSW 2052, Australia. Costello, A.B. & Osborne, J.W. (2005). Best practices in exploratory factor analysis: Four recommendations for getting the most from your analysis. Practical Assessment, Research & Evaluation, 10(7), 1 9. Crocker, L., & Algina, J. (1986). Introduction to classical and modern test theory Orlando, FL: Harcourt Brace Jovanovich. Csikszentmihalyi, M. (1975). Beyond boredom and anxiety San Francisco: Jossey Bass. Dalkey, N. & Helmer, O. (1963). An experimental application of the Delphi method to the use of experts. Management Science 9 (3), 458 467.

PAGE 143

133 Dana, J. D. Jr., (1994). Learning in an equilibrium search model. International Economic Review, 35 (3), 745 771. Daugherty, M. & Funke B. (1998). University faculty and student perceptions of web based instruction. Journal of Distance Education 13 (1) 31 46. Destination Marketing Association International (DMAI, 2006). FAQs about CVBs. Retrieved September 11, 2006, from: http://www.destinationmarketing.org/page.asp?pid=105 Eby D. W., Molnar L. J., Cai L. A. (1999). Content preferences for in vehicle tourist information systems: an emerging tourist information s ource. Journal of Hospitality & Leisure Marketing, 6 41 58. Ewing Taylor, J. (1999). Student attitudes towards web based courses. University of Nevada Reno Retrieved September 11, 2006, from: http://wolfweb.unr.edu/homepage/jacque/research/student_attitudes.html Farquhar, J. D., & Surry, D. W. (1995). Reducing impositions on working memory through instructional strategies. Performance & Instruction, 34 (8), 4 7. Fle ming, M. L. & Levie W. H. (Eds.). (1993). Instructional message design (2 nd ed.). Englewood Cliffs, N.J: Education Technology Publications. Fleming, M. & Levie, W. H. (1995). Instructional message design: Principles from the behavioral and cognitive scien ces Englewood Cliffs, NJ: Educational Technology Publications. Franks, K. (1996). Attitudes of Alaskan distance education students toward media and instruction. The American Journal of Distance Education, 10 (3) 64 68. Frew, A. J. (2000). Information a nd communications technology research in the travel and tourism domain: Perspective and direction. Journal of Travel Research, 39 136 145. Gagn, E. D. (1985). The cognitive psychology of school learning Boston: Little, Brown. Gagne, R. M., Briggs, L. J ., & Wagner, W. W. (1988). Principles of instructional design (3 rd ed.). New York: Holt Reinhart Winston. Gagne, R. W. (1977). The conditions of learning (3 rd ed.). New York: Holt Reinhart Winston Gallarza, M. G., Saura, I. G., & Garcia, H. C. (2002). Des tination image: Towards a conceptual framework. Annals of Tourism, 29(1), 56 78.

PAGE 144

134 Geirland, J. (1996). Go with the flow. Wired Archive Retrieved February, 2002, from: http://www.wired.com/wired/archive/4.09/czik_pr.html Gibson, J. W. (2001). The journe y to cyberspace: Reflections from three online business professors. Advanced Management Journal, 66 (1), 30 34. Global Reach. (2004). Global Internet Statistics Retrieved November, 2005, from http://global reach.biz/globstats/index.php3. Goldfarb, I., & Kondratova, I. (2003, June). Using visual materials to engage learners Proceedings of the ED Media 2003 ACEE World Conference on Educational Media, Hypermedia, and Telecommunications. Honolulu, Hawaii, USA. Govers, R., Jansen Verbeke, M., & Go, F. M. (2000). Virtual tourist destinations: Assessing their communication effectiveness with and through foreign intermediaries. In D. R. Fesenmaier, S. Klein, & D. Buhalis (Eds.), Information and Communication Technologies in Tourism 2000 Proceedings of the I nternational Conference in Barcelona, Spain, 93 103. Gretzel, U., Yuan, Y., & Fesenmaier, D. R. (2000). Preparing for the new economy: Advertising strategies and change in destination marketing organizations. Journal of Travel Research, 39 146 156. Guthr ie, J. T. (1988). Locating information and documents: Examination of a cognitive model. Reading Research Quarterly, 23 178 199. Hall, R. H. (2001). Web based training site design principles: a literature review and synthesis. (165 172). In B. Kahn (Ed.) Web based training Englewood Cliffs, NJ: Educational Technology Publications. Henney, M. (1981). The effect of all capital print versus regular mixed print as displayed on a micro computer screen on reading and speed accuracy. Education Communications a nd Technology Journal, 31 126. Hodkinson, C. & Kiel, G. (2003). Understanding web information search behavior: An exploratory model. Journal of End User Computing, 15 (4), 27 48. Hubbard, A. (1998). What it means to teach online. Computers and Geoscience s, 24 (7) 713 717. Huizingh, E. (2000). The content and design of web sites: an empirical study. Information and Management, 37 123 134. Hull, L. (2004). Accessibility: It's not just for disabilities any more. Interactions, 11 (2), 36 41.

PAGE 145

135 Innovations in D istance Education. (1998). An emerging set of guiding principles and practices for the design and development of distance education The Pennsylvania State University. Retrieved February 1, 2002, from http://www.outreach.psu.edu/de/ide/guiding_principles / Innovations in Distance Education. (1998). An emerging set of guiding principles and practices for the design and development of distance education The Pennsylvania State University. Retrieved February 1, 2002, from http://www.outreach.psu.edu/de/ide/guiding_principles/ Jesitus, J. (1998). On line CVBs lure delegates, book business. Hotel & Motel Management, 213 (6), 15, 67. Jonassen, D. H. & Kirschener, P. A. (1982). Explicit techniques for structuring text. In D. H. Jonassen (Ed.). The technology of text: Principles for structuring, designing, and displaying text, Vol. 2 Englewood Cliffs, NJ: Educational Technology Publications. Jung, T. H., & Butler, R. (2000). The measurem ent of the marketing effectiveness of the internet in the tourism and hospitality industry. In D. R. Fesenmaier, S. Klein, & D. Buhalis (Eds.), Information and Communication Technologies in Tourism 2000 Proceedings of the International Conference in Barc elona, Spain, 460 472. Kedia, B. L. & Bhagat, R. S. (1988). Cultural constraints on transfer of technology across nations: Implications for research in international and comparative management. Academy of Management Review, 13 (4), 559 571. Keller, J. M. & Burkman, E. (1993). Motivation principles. In M. Fleming & W.H. Levie (Eds.) Instructional message design: Principles from the behavioral and cognitive sciences (pp. 3 53) (2nd ed.). Englewood Cliffs, NJ: Educational Technology. Keller, W. (2002). Geogra phic localization of international technology diffusion. The American Economic Review, 92 (1), 120 142. Kiiski, S. & Pohjola, M. (2002). Cross country diffusion of the internet. Information Economics and Policy, 14 297 310. Knowlton, D. S. (2000). A theo retical framework for the online classroom: A defense and delineation of a student centered pedagogy. In R.E. Weiss, D.S. Knowlton & B.W. Speck (Eds.) Principles of Effective Teaching in the Online Classroom 5 14. Kormier, K. A. (1999). Netstorming: Web site development strategies. Strategic Finance Magazine, 80 (11), 64 67.

PAGE 146

136 Kovacs, P., & Rowell, D. (2001). The merging of systems analysis and design principles with web site development: One universitys experience. T.H.E. Journal, 28 (6), 60 65. Kroi P. Kapsammer, E., Prll, B., Retschitzegger, W., & Wagner, R. R. (1997). Online booking on the net: problems and solutions Research Institute for Applied Knowledge Processing (FAW), University of Linz, Austria. Retrieved September 11, 2006, from: ftp://ftp.ifs.uni linz.ac.at/pub/publications/1997/1197.pdf Laesser, C, & Beiger, T. (2004). Information sources for travel decisions: Toward a source process model. Journal of Travel Re search, 42(4), 357 371. Lake, D. (2001) Americans go online for travel information CNN, June 14, 2001. Available: http://www.cnn.com/2001/TECH/internet/06/14/travelers.use.net .idg/index.html Leiner, B. M., Cerf, V. G., Clark, D. D., Kahn, R. E., Kleinro ck, L., Lynch, D. C., Postel, J., Roberts, L. G., & Wolff, S. (2003). A brief history of the Internet Internet Society, retrieved November 2, 2005 from http://www.isoc.org/internet/history/brief.shtml Linacre, J. M. (2002). Judge ratings with forced a greement. Rasch Measurement Transactions, 16 (1), 857 858. Linstone, H. A. & Turoff, M. (2002). The Delphi method: Techniques and applications Retrieved April 12, 2005, from: http://www.is.njit.edu /pubs/delphibook/ Lohr, L. L. (2000). Designing the instructional interface. Computers in Human Behavior, 16 161 182. MacDonald, L., & Caverly, D. C. (2001). Delivering study skills face to face at a distance. Journal of Developmental Education, 24 ( 3), 44 45. MacLennan, R. N. (1993). Interrater reliability with SPSS for Windows 5.0 The American Statistician 47(4), 292 296. Machlis, S. (1997). Profits elude travel sites. Computerworld, 32 (1), 53 54. Mayer, R. (1992). Cognition and instruction: Thei r historic meeting within educational psychology. Journal of Educational Psychology, 84 405 412. Mayer, R. E. & Moreno, R. (2002). Aids to computer based multimedia learning. Learning and Instruction, 12 107 119. McFarland, R. (1995). Ten design poi nts for the human interface to instructional multimedia. THE Journal, 22 (7), 67 69.

PAGE 147

137 McGovern, G., & Norton, R. (2002). Content critical London, England: Pearson. McGovern, G., Norton, R. & ODowd, C. (2002). The web content style guide London, England: P earson. McKenna, J., Avery, R., & Schuchardt, J. (2000). Technology strategies for enhancing learning. Consumer Interests Annual, 46 200 240. Metros, S. E. (2001). Engaging online learners. Lifelong Learning in Europe, 6 (2), 85 95. Meyer, R. (1982). A descriptive model of consumer information search behavior. Marketing Science, 1 (1), 93 121. Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. The Psychological Review, 63, 81 89. Mil ler, M. J. (1996). Being virtual. PC Magazine, 15 (3), 29. Montilva, J. A., Sandia, B., & Barrios, J. (2002). Developing instructional web sites: A software engineering approach. Education and Information Technologies, 7 (3), 201 224. Moorthy, S., Ratchfo rd, B., & Talukdar, D. (1997). Consumer information search revisited: Theory and empirical analysis. Journal of Consumer Research, 23 263 277. Morrison, G. R. & Guenther, P. F. (2000). Designing instruction for learning in electronic classrooms. In R .E. Weiss, D.S. Knowlton & B.W. Speck (Eds.) Principles of Effective Teaching in the Online Classroom 15 22. Motley, L. B. (2000). The 5 elements of web site alchemy. Bank Marketing, 32 (3), 42 43. Murry, J. W., & Hammons, J. O. (1995). Delphi: A vers atile methodology for conducting qualitative research. The Review of Higher Education, 18 (4), 422 436. Naisbitt, J. (1984). Megatrends: Ten new directions transforming our lives New York: Warner. Newman, F., & Scurry, J. (2001). Online technology pushes pedagogy to the forefront. The Chronicle of Higher Education, 47 (44), B7. OneStat (2005). Retrieved September 11, 2006, from http://www.onestat.com/html/aboutus_pressbox31.html

PAGE 148

138 Palmer, A. & McCole, P. (2000). The role of electronic commerce in creating virtual tourism destination marketing organizations. International Journal of Contemporary Hospitality Management, 12(3), 198 204. Pan, B., & Fesenmaier, D. R. (2000). A typology of tou rism related web sites: Its theoretical background and implications. In D. R. Fesenmaier, S. Klein, & D. Buhalis (Eds.), Information and Communication Technologies in Tourism 2000 Proceedings of the International Conference in Barcelona, Spain, 381 396. Petterson, R. (1999, July). Attention: An information design perspective Paper presented at the International Institute for Information Design (IIID), Vienna, Austria. Pettersson, R. (1998, October). Image functions in information design Paper present ed at the 30 th International Conference of the International Visual Literacy Association, The University of Georgia, Athens, GA. Pike, S. (2002). Destination image analysis: A review of 142 papers from 1973 to 2000. Tourism Management 23(5), 541 549 Polin g, C. (1999). Snaring customers on the web: Youd better learn how and fast. New Mexico Business Journal, 23 (5), 42 43. Poulton, E. C., & Brown, C. H. (1968). Rate of comprehension of an existing teleprinter output and of possible alternatives. Journal o f Applied Psychology, 52 16 21. Rankin, R. O. (1989). The development of an illustration design model. Educational Technology Research and Development, 37 (2), 25 46. Raward, R. (2001). Academic library web site design principles: Development of a checkli st. Australian Academic & Research Libraries, 32 (2), 123 136. Rayman Bacchus, L., & Molina, A. (2001). Internet based tourism services: business issues and trends. Fut ures, 33 (7), 589 591. Reigeluth, C. M. (Ed). (1983). Instructional design theories and models: An overview of their current status London: Erlbaum. Rosen, D. E., & Purinton, E. (2002). Web site design: Viewing the web as a cognitive landscape. Journal of Business Research 57, 787 794. Schley, R. (1997). Travel planning online: Virtual vi sits give travelers a head start. The Futurist, 31 (6), 12. Schuman, J. (1996). Dont tangle up your online visitors with web site design. Hudson Valley Business Journal, 7 (15), 13.

PAGE 149

139 Seal, K. C., & Przasnyski, Z. H. (2001). Using the world wide web for t eaching improvement. Computers and Education, 36 33 40. Sherman, L. W. (2000). Postmodern constructivist pedagogy for teaching and learning cooperatively on the web. Cyberpsychology and Behavior, 3 (1), 51 57. Sigala, M., Airey, D., Jones, P., & Lockwoo d, A. (2000). The diffusion and application of multimedia technologies in the tourism and hospitality industries. In D. R. Fesenmaier, S. Klein, & D. Buhalis (Eds.), Information and Communication Technologies in Tourism 2000 Proceedings of the Internatio nal Conference in Barcelona, Spain, 397 407. Simonson, M. R. (1984). Media and persuasive messages. Instructional Innovator 29 (2), 23 24, 48. Snepenger, D., Meged, K., Snelling, & Worral K. (1990). Information search strategies by destination nave to urists. Journal of Travel Research, 29 (1), 13 16. Spiro, R. J., Feltovich, P. J., Jacobson, M. I. & Coulson, R. L. (2002). Cognitive flexibility, constructivism, and hypertext: Random access instruction for advanced knowledge acquisition in ill structure d domains Retrieved September 11, 2006, from: http://phoenix.sce.fct.unl.pt/simposio/Rand_Spiro.htm StatSoft, Inc. (2006). Electronic statistics textbook Tulsa, OK: StatSoft. Availabl e: http://www.statsoft.com/textbook/stathome.html. Stemler, S E. (2004). A comparison of consensus, consistency, and measurement approaches to estimating interrater reliability. Practical Assessment, Research & Evaluation 9(4). Retrieved July 25, 2006 f rom http://PAREonline.net/getvn.asp Sweller, J. & Chandler, P. (1991). Evidence for cognitive load theory. Cognition and Instruction, 8(4), 351 362. Sweller, J. (1994) Cognitive load theory, learning difficulty, and instructional design. Learning and Instr uction, 4 295 312. Sweller, J., Van Merrinboer, J.J.G. & Paas, F.W.C. (1998) Cognitive architecture and instructional design. Educational Psychology Review 10 251 296. Tabbers, H. K. (1999). The effect of presentation mode on cognitive load: Teaching co gnitive skills with text and pictures. Open University of the Netherlands, Heerlen.

PAGE 150

140 Teh, G. P. L., (1999). Assessing student perceptions of internet based online learning environments. International Journal of Instructional Media, 26, 397 402. Thompson, G. (1990). How can correspondence based distance education be improved? Journal of Distance Education, 15(1), 61 70. Tierney, P. (2000). Internet based evaluation of tourism web site effectiveness: Methodological issues and survey results. Journal of T ravel Research, 39 212 219. Tinker, M. A. (1963). The legibility of print Ames, IA: State University Press. Tsai, P., Cambiano, R.L., De Vore, J.B., & Harvey, R.L. (2000). Graduate students perceptions of distance learning. Academic Exchange, Fall 62 66. Turoff, M., Hiltz, S. R., Li, Z., Wang, Y., Cho, H., Yao, X., (2004). Online collaborative learning enhancement through the Delphi method Proceedings of the OZCHI 2004 Conference, November 22 24, University of Wollongong, Australia. Retrieved Octobe r, 2005, from: http://web.njit.edu/~turoff/Papers/ozchi2004.htm Urbina, S. P. (1994). Attention. In R. J. Corsini, Encyclopedia of Psychology (2 nd ed.). New York: John Wiley & Sons. Vogt, C. A., & Pennington Gray, L. (2002, June). Using behavioral segmenta tion to understand the influence of information on welcome center stoppers Paper presented at the annual meeting of the Travel and Tourism Research Association, Arlington, VA. Vogt, C. A. & Fesenmaier, D. R. (1998). Expanding the functional information search model. Annals of Tourism Research, 25(3), 551 578. Weber, K., & Roehl, W. S. (2002). Profiling people searching for and purchasing travel products on the World Wide Web. Journal of Travel Research 37(3), 291 298. Weitl, F., S, C., Kammerl, R., & Freitag, B. (2002). Presenting complex e learning content on the web: A didactical reference model Proceedings of World Conference on E Learning in Corporate, Government, Healthcare and Higher Education, Montreal, Canada. Weston, C., Gandell, T., McAlp ine, L., & Finkelstein, A. (1999). Designing instruction for the context of online learning. The Internet and Higher Education, 2(1), 35 44.

PAGE 151

141 Whalen, S. P. (1997). Assessing flow experiences in highly able adolescent learners (ERIC Document Reproduction Service No. ED 409 382) Wilhelm, W. J. (2001). Alchemy of the oracle: The delphi technique. The Delta Phi Epsilon Journal 43 (1), 6 26. Winne, P. (1995, August). Challenges to the development of self regulated learning Paper presented at the 6 th Europe an Conference for Research on Learning and Instruction. Nijmegen. Witt, G. A. (1981). How to present information that viewers will remember. Instructional Innovator, 26 (9), 37 43. Yacci, M. (2000). Interactivity demystified: A structural definition f or distance education computer based instruction. Educational Technology 40(4), 5 16.

PAGE 152

142 BIOGRAPHICAL SKETCH Charles W. Lane was born in Brooklyn, New York, and grew up in Miami, Florida. He graduated from the University of Florida with a Bachelor o f Science degree in botany in 1994. He studied micropropagation of wetland plant species with Dr. Michael Kane and graduated with a Master of Science in environmental horticulture from the University of Florida in 1999. His masters thesis was entitled Tis sue Culture Propagation of Sagittaria latifolia After graduation with a Doctor of Philosophy degree in tourism, recreation and sport management, he intends to pursue a career in higher education.