A COMPARISON OF HOME BUYERS' PREFERENCES
REGARDING COST AND NON-COST SPECIFIED OPTIONAL FEATURES











By

LUDWIG MORTIMER DYSON, JR.














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




UNIVERSITY OF FLORIDA

1978














ACKNOWLEDGMENTS


This study represents the contribution of many people.

Dr. Halbert C. Smith, Chairman of the Dissertation Committee, contri-

buted guidance and untiring support in the writing and editing of this

study. Dr. Brisbane Brown provided invaluable support in the develop-

ment of the visual display kit that was used in the personal interviews.

Dr. Wayne Archer also provided useful ideas regarding marketing research

in housing. Special thanks also go to Dr. Louis Gaitanis and

Dr. Clayton Curtis for their suggestions and support.

Mr. Forrest Hope provided the local cost data that were necessary

for the success of this study. Special thanks go to Alex Morrison and

David Consbruck of the School of Building Construction for their help

in conducting the personal interviews and the construction of the visual

display kit. Tim Breen provided assistance in the statistical analysis

and computer programming. Special thanks also go to Carol Riggall for

her help in the mail-out questionnaires and other invaluable support.

I would also like to thank my family for their guidance, support,

and patience during this study. I owe to them the credit for the many

good things which have come to me.

Last, I would like to thank Phillip Pickens, MAI, SREA, for his

guidance during my graduate studies. It was through his efforts that

my enthusiasm for real estate was transformed into an ordered study.















TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS ....................................... ....... ii

ABSTRACT .................................................. .. v

CHAPTER

I INTRODUCTION ........................................ 1

Statement of the Problem ............................. 1
Setting of the Study ................................. 2
Study Design ......................................... 2
Organization of the Study ............................ 3

II A REVIEW OF RELATED LITERATURE ....................... 5

Feature Preferences Studies .......................... 11
Home Buyer Decision-Making Studies ................... 22
Consumer Durable Studies ............................. 24

III DESIGN OF THE EXPERIMENT ............................. 27

Hypothesis .......................................... 27
Population ........................................... 27
Sample ............................................... 28
Instruments .......................................... 29
Collection of Data ................................... 30
Analysis of Data ..................................... 31

IV PRESENTATION AND ANALYSIS OF DATA .................... 33

Summary of Individual Differences .................... 61
Test of Differences Between Sample Means ........... 65
Test of Differences Between Homeowners'
and Apartment Dwellers' Preferences ............... 67
Optional Feature Tradeoffs ........................... 75
Demographic Characteristics .......................... 77

V CONCLUSIONS AND RECOMMENDATIONS ...................... 80

Conclusions .......................................... 80
Recommendations ...................................... 84











APPENDICES

A LETTERS TO HOMEOWNERS AND APARTMENT DWELLERS ...... 87

B SURVEY FORM TO ACCOMPANY COST DATA AND
GRAPHIC DISPLAY ................................ 90

C MAIL-OUT QUESTIONNAIRES ........................... 102

D PHOTOGRAPHS OF GRAPHIC DISPLAY KIT ................ 127

BIBLIOGRAPHY ................................................. 142

BIOGRAPHICAL SKETCH ................................ ... ...... 146














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


A COMPARISON OF HOME BUYERS' PREFERENCES
REGARDING COST AND NON-COST SPECIFIED OPTIONAL FEATURES

By

Ludwig Mortimer Dyson, Jr.

December 1978

Chairman: Halbert C. Smith
Major Department: Real Estate

The major objective of the study is to determine whether the

presence or absence of price in the decision-making process of home

buyers produces different preferences for optional features. To

accomplish this objective an improved methodology for surveying

potential home buyers was developed. The previously employed

methodology allows consumers to choose from a list of features that

may not be feasible within their budgetary constraints. This study

used both methodologies to determine whether significantly different

preferences are obtained from each procedure.

The hypothesis tested is that there is no difference between the

preferences of consumers who are made aware of the price of each optional

feature and those who are not informed as to the price of each feature.

Using mail-out questionnaires without price information regarding

optional features, preferences were ascertained for one group of con-

sumers. A second group was personally interviewed using a graphic









display kit showing each optional feature accompanied by its price.

In both surveys consumers were constrained by a maximum price of

$38,500.

Responses were obtained from a sample of 120 home owners and

apartment dwellers in the Gainesville, Florida, area. Chi-square

testing was used to determine whether individual differences exist

between the two groups surveyed. These tests indicated that 16

individual cost related preferences were significantly different

between the two groups. This result indicates that the group informed

about the prices of optional features produced different preferences

than the uniformed group. An additional test of differences of

sample means indicated that the two groups do not have the same overall

preferences.

Further Chi-square testing identified individual preference

differences between previous homeowners and those who have never

owned a hone. Additional testing determined that no significant

level of difference exists between apartment dwellers and homeowners.

The hypothesis of no differences between a consumer group that

was informed about prices and one not so informed was rejected at

the 0.05 level of significance. A corollary conclusion is that a

survey using pictorial and graphic descriptions accompanied by price

data produces a more accurate assessment of consumers' preferences

than previously used methods of opinion surveys.














CHAPTER I

INTRODUCTION


The purchase of a home by an American family often represents the

single largest purchase made in a lifetime; therefore, consumers in

the housing market are concerned that they make the best choice in the

purchase of a home. This decision-making process of the home buyers

is a complex process. For one to try to describe the process would

lead to incomplete results.

One important consideration essential in the purchase decision is

the price of the home, however. Consumers are generally constrained

in the home purchase by family income. In recent years, the increasing

cost of housing has created the need for cost reduction techniques, if

the average American family is to be able to purchase a conventionally

built, single-family home.


Statement of the Problem


The builder and developer must be willing to serve specific

markets; they must understand market segmentation. Builders, after

identifying a specific market, must respond to its needs and desires.

Identification of needs and preferences is often a difficult task.

Caution must be exercised in the attempt to ascertain consumers'

preferences in housing. No simple questionnaire or check list compiled

as a side-line by the architect or builder will furnish the information

needed. The potential consumer will check all items vaguely conceived

1









as desirable, yielding a "Christmas list" that builders will rarely be

able to supply. Various features of home construction must be appraised

in relation to their costs [Riemer, 1951].

While a few studies by housing publication companies, building

associations, and home building manufacturers have attempted to deal

with consumer motivations, needs, attitudes, and preferences, no research

has attempted to acquaint consumers with the price of each optional

feature offered. When confronted with a budgetary constraint of a

maximum affordable price, consumers could presumably optimize the hous-

ing space, optional features, and design in accordance with their

preferences.


Setting of the Study


This study was conducted in the Gainesville, Florida, metropolitan

area. Using random sampling, subjects were chosen from families living

in subdivisions containing houses in the general price range of the

study ($38,500) platted since 1975 and apartments in the Gainesville,

Florida, area. No contention is made that the responses of the

subjects from this area are typical for consumers throughout the

nation. The tests of preference differences could be replicated in

other areas substituting local cost data from a specific area.


Study Design

One of the major tasks of this study was to develop a survey

instrument to solicit responses of potential home buyers. To

accomplish this, a graphic display kit was designed and constructed.

The function of the graphic display kit was to aid respondents in









making choices by showing size perspective of rooms, layout arrange-

ments, optional features, and the price of each feature. The kit was

designed in flip-chart form and packaged so that it could be trans-

ported easily by the interviewer and set up in the respondents' homes.

The price of each optional feature was obtained from contractors'

detailed cost analyses of comparable homes included in this study.

By using this "hybrid" form of cost comparison, each feature and

component of the structure (home) was individually priced. This

allowed the respondent to see each choice as well as to be aware

of the price of each optional feature (see Appendix D).

The kit was combined with a survey form on which the interviewer

recorded all responses made by the subjects. Cumulative costs and

square footage were also recorded on this form.


Organization of the Study

Chapter II contains a review of the literature related to market

research in housing. Specific emphasis is placed on feature preference

studies. The majority of these studies were completed by building

trades journals, building industry organizations, and real estate

development firms. The techniques used in these studies and the

results are reviewed to identify potential weaknesses, as well as the

contributions, made by the research.

In Chapter III, the design of the experiment is discussed.

Sampling techniques, data gathering techniques, instrument construction,

and statistical tests employed are outlined in this section.






4


Chapter IV presents the data and the statistical tests that were

performed to test the hypothesis of the study. Chi-square tests were

used to test for differences of individual preferences, and a test of

differences of sample means was used for overall differences in

preferences. Tradeoff information and demographic characteristics are

presented in this section also.

Chapter V summarizes the findings and contributions of the study.

Recommendations for additional research are included.














CHAPTER II

A REVIEW OF RELATED LITERATURE


This chapter reviews studies related to feature preferences of

consumers in the housing market. The majority of research in this area

has been done by building industry-related research firms and pub-

lishing firms of building periodicals. The review of these studies

seeks to identify the methodology employed and a summary of conculsions

found by these researchers.

In general, marketing research plays an advisory role in the over-

all marketing process. The information acquired and the analysis

of the findings are used to make recommendations to the firm's

management as to appropriate action to be taken in the marketing

of a product [Tull and Hawkins, 1976]. Producers in the housing

industry have a very complex product to sell, and knowledge of con-

sumers' preferences is essential for a successful marketing program.

By observing these preferences in a systematic way, designers and

builders may obtain data upon which to design and build homes that are

indicative of these preferences.

Housing researchers are often disenchanted by builders' claims

that consumers of housing do not buy the type or size of home they

claim to prefer. In one study [Whipple, 1971] this paradox is explained

as constrained maximization. People cannot afford what they prefer

(no feasible home at their budgetary constraint exists), so they must

reformulate their preferences subject to new constraints.

5









Prediction of housing choice is difficult for housing market

researchers because buyers purchase houses that are significantly

different from the houses they claim to prefer. This is due, in part,

to the lack of meaningful research in housing that allows the consumers

to indicate preferences that are within their budgetary constraints

[Reimer, 1951].

The need for more fruitful research is echoed throughout building

journals. Lester Goodman, Vice President of Marketing of D. Lusk and

Son, offers this observation:

In most industries, countless management hours are spent
researching, analyzing and interpreting consumer actions and
motivations. The builder, however, budgets little or nothing
on consumer research and then compounds this error by placing
poorly trained, ill-equipped salesmen at the sales site to
handle this complex situation. [Goodkin, 1975, p. 56]

Some successful building firms have completed market research

studies. In a study of builders in twelve cities, conducted by the

Director of Marketing Information Network of Professional Builder and

Apartment Business, two conclusions were formulated [Goodkin, 1975].

First, the most successful builders in these cities had thoroughly

researched their markets. They were aware of the consumers' wants in

their market and responded to the market by providing a product at a

price that was compatible with the financial abilities of the consumers.

In these cities, the other group of builders, unsuccessful in their

developments, was not aware of the preferences of the consumers in the

market. In other words, this group of builders was guilty of designing

and building houses based on their own preferences, instead of those

of the consumers.









Secondly, the study concluded that in addition to responding to

consumers' preferences, the successful builders were confident of the

product and followed up with an aggressive marketing program. However,

it was noted that a marketing program, no matter how aggressive, will

not succeed if the product does not meet the needs of the consumers

purchasing the house [Birkner, 1975].

Historically, successful housing developers have focused on two

basic ingredients: location and financing. If the developer could

purchase a parcel of land at the "right" price and create an attractive

financial package, success was all but guaranteed. The kind of

housing, or the type of ultimate consumer, was only secondary con-

sideration.

Builders must ask the same questions as other producers of

consumer goods:

(1) who are the potential buyers; and

(2) what do they want [Tull and Hawkins, 1976].

Instead of believing they are experts about consumers wants and pre-

ferences, builders must seek to identify consumers' preferences.

A number of changes occurring in the housing industry are con-

vincing developers that intuition is not enough in today's market:

(1) Developments are getting larger, so more units have to be
marketed.

(2) More dollars are involved, so the risk is greater.

(3) Producers are creating more than one development simul-
taneously, stretching management intuition (and risk)
over several developments at once.

(4) Producers are going public, and it is easier to persuade
investors on the basis of market research than on the
basis of intuition [Hardin, 1972].









Until recently, very few people in the building industry recog-

nized the need for housing research. Even the government sector, in

spending vast sums of money for home purchasing programs, has done

only minimal housing research. W. L. Wheaton, in testimony before a

sub-committee of the Committee on Banking and Currency of the U. S.

House of Representatives, pointed out that the U. S. Department of

Agriculture spends more money on housing research for 6 million

farm families than the other agencies spend on research for nonfarm

families [Kelly, 1960].

As Kelly points out, the average builder

. seems to think a great deal harder about the details of his
design after it has been adopted and was going into production or
even into the marketing phase than he had in the first place about
the broad principles upon which the design was based. This is
perhaps understandable since time, energy, and money for broad
analysis are often very limited once operations are under way,
while in the early stages of organization and design, the problems
of financing and creating a production and distribution system
seem very small and remote. When the realization comes that the
first step in the pattern of operations (design) should have been
altered in order to better perform the last (particularly selling
the house), the die has been cast. [Kelly, 1960, p. 236]

Proceeding from the accepted opinion that research in housing is

necessary, certain acceptable techniques have been advocated to gain

the needed information to formulate consumer preferences. Four

techniques that are generally employed are opinion surveys, pre-testing

of the product, test living, and post-testing the house design.


Opinion Surveys

This technique allows the researcher to ascertain specific

feature preferences. The use of both direct questions and indirect

interviewing can achieve significant results. The use of direct

questions allows the interviewer to ask very specific questions









regarding preferences. For example, the consumers may be asked

whether they would like a patio in the backyard. Indirect inter-

viewing uses likes and dislikes of features that consumers have seen

recently in other houses as the basis for design. The use panels of

consumers reveals an indication of any distinct pattern of consumer

preferences within a market segment. Such panels allow group dis-

cussion of similar preferences. Group discussions may often elicit

responses available from no other technique, provided the participants

are unconstrained [Kelly, 1960].


Pretest the Product

Modern consumer researchers and marketers know the value of

introducing the new product to a limited market before going to the

entire market with the new product. Although houses cannot be tested

in exactly the same way, a hybrid form of test marketing can be done

in an exhibit house. Using this technique, trained observers are

stationed in sample houses, seeking to record the remarks of potential

consumers. In addition, the trained interviewers may question a

cross-section of viewers about specific design features.

Another variation uses a carefully selected sample of typical

American families. These families are shown specimen housing using

different design features and asked to give their reactions after-

wards. Using personal or mail questionnaires, respondents are

questioned on general reactions to house styles, construction features,

and specific design features. Demographic information such as

occupation, income, family size, and educational background can be

gathered to discern differences in preferences among different market

segments.









The pretest method was employed by a large builder in Chicago.

Real Estate Research Company of Chicago supervised the marketing that

obtained data on 58 classifications of structures, basic style, wall

structure, orientation of lot, window types and room arrangements.

The alternatives were integrated into 2U specimen houses that were

actually constructed. The information received from each respondent

was cross tabulated to reveal style trends, preferences for specific

kinds of materials, and popular floor plans [Kelly, 1960].


Test Living

The method allows families to live in a test home, usually rent

free, for an extended period of time. The responses about the features

they liked or disliked are recorded, as well as suggested improvements.

Careful selection of families reflecting occupation, income, family

size, and other demographic characteristics is essential, if market

segments are to be identified [Kelly, 1960].


Post-Test House Design

This technique involves interviewing owners of houses to determine

their likes and dislikes of amenities and design features. The

researcher is in a position to advocate design modifications if research

warrants such changes [Kelly, 1960].

The basic format of virtually all housing research has been some

form of opinion surveys. Using both direct and indirect interviews,

the researchers have attempted to ascertain consumer attitudes and

preferences related to a single-family housing. The following studies

are indicative of methodology employed and results obtained.









Feature Preference Studies

National Consumer Survey on Housing 1975

In the first comprehensive study conducted by the Bureau of

Building Research for Professional Builder and Apartment Business,

questionnaires were mailed to 10u3 potential home buyers

[National Consumer Survey on Housing, 1975]. All indicated they

planned to purchase a new home in the immediate future. From the

original sample, 8U7 responded, indicating their attidues and

preferences regarding housing.

The study was composed of two major sections. In Section I, the

researchers sought to identify general consumer attitudes toward

housing. They found that in excess of 90 percent of all families

preferred the typical single-family home and that approximately 75

percent wanted homes costing less than $45,000.

In Section II, the study sought to segment the market by region

and typical demographic characteristics including income, education

age, and family size. However, several inconsistencies appear when

some of the preferences and attitudes are considered using the demo-

graphic characteristic of income. When comparing income of the

respondents to the price range of the home they planned to purchase,

there seems to be sufficient evidence to say that the respondents are

not realistically stating demand. It appears that in the study, desire

is often confused with demand. That is, desire is what they want and

often not backed with the income necessary to create demand.

Further unrealistic indications of preferences were noted in the

selection related to recreational facilities within the development.

For example, in the price ranoe of "Under 25,000," 70 percent of the









respondents indicated that it was important that the development have

swimming facilities, 60 percent wanted tennis courts, and 50 percent

wanted sauna and gym facilities. Again, while these are desired

amenities for an ideal development, a developer could not provide these

amenities to purchasers in the "Under 25,000" price range of home buyers.

When asked how many bedrooms and baths the home buyers preferred,

they responded that three or four bedrooms and two baths were desired.

Again the economic feasibility must be questioned. While it may have

been possible to supply a home with three bedrooms and one bath for

under $25,000 in 1975, it is hardly conceivable that four bedrooms and

two baths were possible. Yet, 20 percent wanted two baths in the

"Under 25,000" category.

Several of the questions required the respondents to rank the

order of importance of various feature options. One question required

the respondent to rank each feature option (some 29 options) on a scale

from one to four with one being least important and four being most

important. The rankings are presented using a mean ranking (no standard

deviation). In the lower price category it is almost impossible to

determine a difference in importance of some 12 options because the

means are so close.

The researchers attempted to remind the respondents to indicate

preferences realistically according to the price range they planned to

purchase by inserting the phrase "keeping in mind your home's price

range what type of . "; however, as shown in the foregoing examples,

many choices were not related to the price range indicated.









National Consumer Survey on Housing 1976

In the second study commissioned by Professional Builder and

Apartment Business, the sample size was reduced to 400 with 327 con-

sumers returning the mail-out questionnaire [National Consumer Survey

on Housing, 1976]. This study in many ways resembled the original

survey completed in 1975. Many of the same problems noted in the 1975

study were also found in the 1976 study.

The researchers sought to "fine tune" some of the areas covered in

the first survey by asking more specific questions. It was during this

period of time that housing costs were escalating at a rapid rate and

builders were challenged to produce a marketable home while reducing

costs. Consumers were asked whether they would accept a smaller home

as they had accepted the compact car. Over 40 percent of the respondents

indicated that reducing the size of the home was not an acceptable means

of curbing cost increases in housing. But, where asked for specific

responses to means of reducing operating costs, over 80 percent indicated

that they would be willing to spend an additional $600 at the time of

initial construction, if it would mean a saving of $100 per year in

operating expenses.

The next series of questions was designed to solicit responses

regarding alternative ways to reduce initial home construction costs.

An average of 80 percent indicated that builders should design and build

homes that could be expanded at a future time. This indicates that while

many consumers were not willing to accept a smaller home, they would

consider a smaller home if it was designed and constructed for expansion

at a future time. Only 8 percent indicated they would accept a









smaller lot and 12 percent indicated they would give up a garage for a

carport as cost saving measures.

In specific preferences for kitchens and baths, a large majority

wanted the standard all-wood cabinets and cast iron tubs with ceramic

tile. It appears that consumers in this study were not willing to

accept cost saving amenities such as plastic laminated kitchen cabinets

or prefabricated fiberglass shower/tub assemblies.

Again, in this study energy saving ideas such as heat pumps and

various solar items received high responses. However, it should be noted

that even in the lower price ranges of homes, many respondents indicated

they would be willing to spend an additional $6,000 for solar heating,

even though the maximum amount they planned to spend for housing was

$25,000.


National Consumer Survey on Housing 1977

The third study conducted for Professional Builder and Apartment

Business used a sample size of 769 and had a response rate of over

80 percent based on 633 questionnaires returned [National Consumer

Survey on Housing, 1977]. Much of the same type of information sought

in the first two studies was again solicited in this survey; however,

the results were different in several areas.

It appears that because of constantly increasing costs of new homes

from 1975 to 1977, consumers were more realistic in their choices of

homes and amenities preferred. It is thought that the lower range of

homes (under $35,000) began to lose appeal to consumers because of

unavailability. Although some respondents still indicated prefer-

ences for homes in this range, there was a significant decrease.








Another indication that inflation and rising housing costs were a

significant factor was shown in the section related to delayed pur-

chases as a result of the loss of purchasing power. In the 1976 study

when asked what areas would consumers delay purchases, fewer than two

percent indicated housing; yet in the 1977 survey, over 10 percent said

they would delay housing purchases. In this study, as in the two pre-

vious studies, consumers still seemed to be unwilling to accept a

smaller home. The percentage of respondents actually decreased by

approximately 10 percent. It appears that consumers also changed their

preferences in the alternative ways to reduce costs of a new home.

In the previous two studies respondents considered expandable design to

be the appropriate way to achieve lower cost, indicating that they

planned to add on at some time in the future. In this study a signifi-

cant number of respondents (40 percent) indicated that they would be

willing to accept standard designs that would reduce costs because of

uniformity of construction.

An additional section was added to this study dealing with payment

abilities of the potential home buyers. The respondents were asked to

indicate the maximum monthly payment they could afford including

principal, interest, real estate taxes, and homeowner's insurance.

Using cross tabulation, the researchers showed the price range of the

home chosen and the responses of the homebuyers with regard to the

maximum payment they could afford. In many instances the price of the

home was far in excess of the amount of the monthly payment they could

support. For example, almost 20 percent of the respondents who wanted

a mean price home of $35,000 could afford a monthly payment of less

than $200. In another case, almost 25 percent of the buyers desiring









a mean priced home of $40,000 could only afford a total monthly payment

between $200 and $250. Using mortgage terms and interest rates of 1977,

it can be determined that a significant percentage of the respondents

desiring a given priced home could not even afford interest and

amortization, much less real estate taxes and insurance.


National Consumer Survey on Housing 1978

In the most recent of the four housing studies completed, a new

survey technique was used [National Consumer Survey on Housing, 1978].

In addition to the techniques used previously that employed a large

sample (849 questionnaires were mailed out, 663 were returned), builders

were also surveyed. The builders surveyed were randomly selected from

the circulation list of the sponsoring organization, Professional Builder

and Apartment Business. Some 3,000 builders were mailed a questionnaire

and 678 responded, yielding a 23 percent response rate. The builders

were surveyed to determine whether they were building single-family homes

in accordance with the preferences indicated by the respondents in the

consumer section of the study.

One noticeable difference appeared in the first section of this

study. In all three previous studies, a question was to determine the

price range of homes being sought by potential home buyers. The minimum

category in the first three studies was "Under 25,000". In this study

the minimum category was "under $35,000," indicating that consumers were

aware of the rapid increase in housing prices. Also, in this study more

detailed questions were used to ascertain preferences regarding various

energy saving materials and devices.

As in previous studies, consumers still seemed to want the typical

home with three or four bedrooms, two baths, double car garage, a large








entry, separate dining area, and a large kitchen. In addition to these

items preferred, consumers seemed to expect more for the price of the

home. This is supported by the responses received when asked, "Keeping

in mind the price range of the home you indicated, which of the following

features do you feel should be included as standard items?" Over 50

percent of the respondents in the "under $35,000" category felt that a

range hood with fan, stove, dishwasher, disposal, refrigerator, washer,

and dryer should be included as standard items. In addition over 50

percent of this same category of home buyers felt that a fireplace,

insulated windows, smoke/fire detectors, heavy duty security locks,

storm windows and screens, central air conditioning, and power attic

ventilator should be included as standard items in the price of this

"under $35,000" home.

In an effort to determine the respondents' abilities to purchase

the housing indicated as their preference, the researchers also asked

the consumers to indicate the maximum monthly payment they could afford.

Some 15 percent of the respondents indicated housing choice in the range

from $35,000 to $65,000, yet stated that the maximum amount they could

pay each month was less than $200. This evidence seems to indicate that

many of the respondents' preferences were not indicative of their ability

to pay, but were only indications of desire.

The unique section of this study reports a survey of builders to

determine the interaction of supply with demand as evidenced by consumer

preferences. To determine what was being built, the researchers asked

the builders to indicate the price range and various feature options

offered on their best selling homes.








In the demand section of the study, 50 percent of the consumers

indicated that they intended to purchase a home costing $50,000 or less;

however, in the builder's section contractors responding indicated that

fewer than 35 percent of the homes being constructed sold for $50,000 or

less. Similarly, 40 percent of the consumers in the "under $35,000"

category expected the house to have a two-car garage, while fewer than 20

percent of the builders included this type of garage in the homes being

constructed in this price range.

In the section of the study dealing with cost saving alternatives,

the researchers sought to determine the items homeowners would be willing

to sacrifice, thus lowering the cost of a new home. Fewer than 10 per-

cent of the respondents indicated that they thought reducing the lot size

would be an acceptable means of lowering costs; however, 40 percent of

the builders disclosed that this idea was a dominant technique in their

efforts to control costs. Likewise, fewer than 20 percent of the

respondents expressed the belief that building a smaller home seemed to

be a favorable alternative in cost reduction; however, more than 50

percent of the builders specified a reduction in square footage as a

measure used to reduce construction costs effectively.

The opposing viewpoints held by consumers and builders should not

go unnoticed. In a statement regarding changing consumer attitudes,

M. R. Robinson, Director of Bureau of Building Marketing Research, cites

lot size reduction as one area in which consumers are willing to sacri-

fice to assure lower costs [Robinson, 1974]. Robinson further states

that single-family attached housing will be as acceptable as the conven-

tional single-family detached home; although four comprehensive studies









by his organization indicated that over 95 percent of consumers below

60 years of age would not prefer attached single-family houses.


1977 Consumer Attitude Study

This study, commissioned by the Orange County Chapter of Building

Industry Association of California, differs significantly from the four

previously reviewed studies [Consumer Attitude Study, 1977]. Based on

1,100 on-site interviews with potential new home buyers at 80 different

residential developments in five counties surrounding Los Angeles, the

study sought to determine consumer preferences and attitudes in almost

all areas of housing.

By the admission of the editor, "no effort was made to statistically

pre-determine an 'average' price range or design to make up the sample."

Furthermore, the researchers admitted that the consumers attracted to the

subdivisions used in the survey were "undoubtedly influenced in their

responses by the price range and design of the product they had just

visited." Thus, the results of the study were subject to "cause effect"

distortions.

One of the primary goals of the study was market segmentation. The

researchers identified 14 separate market groups and listed design impli-

cations for each segment. By this identification method, the study was

to provide detailed information about each group, thus enabling builders

to respond to the demands of their particular market group.

A major part of the study was devoted to consumer attitudes related

to energy saving products. The researchers found that 17 percent of the

respondents would not pay an additional $500 for insulated windows and

30 percent indicated "negative attitudes" to paying the additional

dollars. The builders, using the results of the study as a guide in









responding to market demands, were told that marketing must play an

"educational" role when insulated windows are included as a standard

or optional item. Based on the results of the survey, builders are

told that with "proper" selling techniques, approximately 35 to 64

percent of their customers could be expected to purchase insulated

windows.

The objective of the study was to ascertain consumer preferences

and attitudes. Negative responses of home buyers, as stated earlier,

are often indicators of preference and attitudes. The major negative

comments of home buyers regarding the developments visited were

(1) price too high

(2) houses too small

(3) poor layout or floorplans

(4) lots too small

(5) bedrooms too small

It appears that the first two comments made by consumers were made

because of the lack of knowledge of building costs; however, the

remaining negative remarks can be researched systematically in the

market, and homes can be constructed responding to preferences of

potential home buyers.

One area addressed in this study that may have been overlooked in

the previously reviewed studies concerns the substitutability of used

housing. As shown, market analysis should also include demand analysis.

Researchers found that almost one-half of the respondents considered

"resale" homes to be an alternative, particularly if the sale price is

substantially less than that of new homes.









An attempt was made to determine the order of importance of various

amenities and feature options to be omitted, if it was necessary to

reduce the purchase price. It was determined that consumers are willing

to forego recreational amenities within the development, if the purchase

price must be reduced. Fewer than 10 percent of the home buyers would

give up the garage or reduce the lot size by 10 percent to reduce the

purchase price.

The remainder of the survey attempted to determine preferences and

attitudes regarding such items as number of bedrooms, baths, and

garages, interior finish, exterior design, and neighborhood qualities.

Using various demographic characteristics, the researchers segmented

the markets according to groups showing the differences in preferences.

Although this study was completed for a specific region (Southern

California), the results are similar to national studies reviewed

previously. The methodology employed is questionable because of the

suggestive effect caused by the recent exposure to new homes during the

consumers' visits to the residential developments; however, the technique

is used by other research groups, as evidenced by the following study.


Consumer Preference Survey Insight 1977

This study was also conducted at subdivisions throughout Southern

California, and the sample of almost 700 consumers represented a con-

venience sample [Consumer Preference Study, 1977]. The respondents were

potential home buyers visiting the 16 subdivisions of both detached and

attached single-family homes. The price range of the homes in the

development was from $32,000 to $120,000.









To justify the validity of the preferences of consumers, the

researchers stated that buyers were willing to "stretch" their incomes

to purchase higher priced housing. A significant number of respondents'

"demand" for housing priced in excess of the typical 2.5 times annual

income was noted. The researchers indicated that reliance upon existing

equity was the reason many potential buyers were demanding housing

priced at almost three times annual income; however, demographics

indicated that one-third of the potential buyers were first-time buyers,

hence no equity.

As in other studies, only a small percentage of the consumers

(4.5 percent) were willing to reduce lot size, even if it meant saving

$1,400; however, almost 60 percent of the consumers would pay from

$1,400 to $9,800 more to obtain a larger lot.

The survey instrument used in this study used perspective render-

ings, elevations, and floor plans to graphically illustrate various

alternatives. It appears that more specific responses were generated

using this technique. The researchers were able to determine feature

preferences regarding master bedroom placement, kichen/family room

arrangement, and dining area placement.

An attempt to include the prices of some features and amenities was

noted; however, because of the extreme differences in the purchase price

of the homes in the various developments, consumers had an almost

inexhaustible list of choices available.

Home Buyer Decision-Making Studies


In studies conducted by persons not directly associated with the

home-building industry, researchers have attempted to ascertain the








dominant decision-maker in feature preference selection in home buying

[Munsinger, et al., 1975, Davis and Rigaux, 1974 and Hempel, 1974].

The primary objective of these studies was to determine whether the hus-

band or wife was responsible for making various decisions or were the

decisions jointly made.

One study [Hempel, 1974] identified five important elements

regarding the purchase decision. These elements were tested to deter-

mine perceived dominance by one spouse or whether the decision was

jointly made. Based on data obtained from families in the Hartford,

Connecticut, area, it was found that the decision to move to another home

was considered to be a joint decision; however, it was found that the

husband plays a dominant role in choosing the source of financing for

the new home.

The Hempel study also identified conflicts in the husband-wife

combined decision-making process. The five major areas of conflict

identified were location, feature options (e.g., fireplace, garage),

price, architectural style, and floor plan; however, the researcher

found that 62 percent of the conflicts in the decision process were

resolved in favor of the wife [Hempel, 1974].

Overall, studies have found that decisions regarding housing are

jointly made; however, in studies that subdivided the purchase decision

into several interrelated decisions, evidence was found that indicated

significant variability in the relative involvement of husband and wife.

The husband's influence was highest for decisions regarding price range

and the decision to relocate, while the wives' influence was highest

in deciding the number of bedrooms, colors, and other house features

[Bernhardt, 1974 and Davis and Rigaux, 1974].









Consumer Durable Studies


Contrasting the level of sophistication used in marketing research

in the housing area with that used in other consumer durables, the

writer noted a significant difference; furthermore, the writer sought to

examine techniques employed in studies of other consumer durables with

regard to their applicability to housing market research.

The automobile purchase, representing the second most important

decision (in terms of money) made by consumers, provided excellent

application to housing research. In considering which automobile to

purchase, the consumer faces many of the same pre-purchase decisions as

the home buyer. For example, because of the concern for increasing

energy costs, both consumer groups must decide between various alterna-

tives as to the energy consumption thriftness of both automobiles and

homes. Consumers in the automobile and housing markets are usually

subject to a price constraint and must choose between feature options

available.

Market segmentation, using various demographic characteristics,

social, attitudinal, and personality variables, has provided researchers

with evidence of target markets for different automobiles and options

supplied by the manufacturer. Recent innovations in the automobile,

such as the rotary engine Mazda, have provided researchers an excellent

laboratory to segment buyers who are willing to "venture" into newly

developed products [Feldman and Armstrong, 1975].

The "Mazda study" identified age and education to be important

variables related to the acceptance of innovative products. The








noninnovators were found to have higher education, while the younger

buyers, regardless of sex, proved to be the innovators.

Other studies seek to use multiple regression techniques to

determine the usefulness of demographic characteristics in predicting

consumer choice in the purchase of an automobile [Wiseman, 1971 and

Peters, 1970b]. The researchers found that for each classification of

automobiles, demographic characteristics of buyer groups were evident,

and the demographic variables identified in each case explained a high

percentage of variation.

Researchers are constantly attempting to use income and occupation

as explanatory variables in the automobile market. By combining the

two variables, research efforts have sought to identify purchase

behavior based on occupational classification income [Peters, 1970a].

Using total family income and the median income for a given occupational

classification, the family was classified as average, "underprivileged,"

or overprivilegedd." The significant conclusion of the study was that

the overprivilegedd" of an occupational classification tended to have

the same buying behavior as the next higher occupational classification.

The study points to the potentially valuable application to other con-

sumer durable markets.

Market studies of other consumer goods were reviewed. These

studies included information effects on the pre-purchase decision,

brand switching in consumer durables, taste tests for food products,

and store preference tests for location analysis. The methodologies

employed in these studies were not relevant to this study.

Although some comprehensive studies have been completed to deter-

mine consumer preferences in housing amenities and feature preferences,






26


little attempt has been made to verify these preferences. Consumers

have been given an almost inexhaustible list to choose from, with little

regard for their ability to afford the choices made. Although pre-

ferences may be ascertained by asking consumers for their choices of

housing related amenities and feature options, some thought must be given

to the price of these choices. This study is designed to extend the

methodology and literature to include research on the effect of price

on consumer preferences; additionally the study will introduce the trade-

off element when the consumer reaches the maximum affordable price.















CHAPTER III

DESIGN OF THE EXPERIMENT


Hypothesis


This study was designed to compare responses of consumers in the

housing market using different methodologies. As shown in Chapter II,

one method of ascertaining consumer preferences is accomplished using

mail-out questionnaires. In a few questions, the price of an option

or feature was included as a basis for consumer choice. The methodology

developed in this study informs the respondents of the price for each

choice made. The specific hypothesis was

There is no significant difference in the preferences of home
buyers who are aware of the price of each optional feature and
home buyers who are not aware of the price.

Responses derived from the sample served as a proxy for consumer prefer-

ences of home buyers.


Population


Two populations served as the base for this study. One population

was composed of all families living in detached single-family housing in

the general price range of the study ($38,500) in Gainesville, Florida.

Additionally, the population was confined to those families living in

homes in subdivisions platted since 1975. Although the selection of

1975 was arbitrary, it insured that all families were living in homes

that were constructed using materials and current building technology.









The second population was families living in rental housing in

Gainesville, Florida, that did not own a single-family home previously.

It was from those two strata that the samples were drawn.


Sample


A basic requirement of using Chi-square testing is that the sample

size must be large enough so that each cell in the contingency table

contains at least five observations; therefore, it was determined that

a sample size of approximately 30 would provide sufficient responses

for the test. Since two different survey instruments were used and

two consumer groups (homeowners and apartment dwellers) were surveyed,

a total of approximately 120 subjects were included in the samples.


Homeowners

Plats of subdivisions containing homes in the general price range

of the study were obtained. Using lot numbers of each subdivision and

a random numbers table, a sample was drawn. A pretest of subjects

indicated a response rate of approximately 80 percent for personal

interviews and 30 percent for mail-out questionnaires could be expected.

Thus 40 families were selected randomly to obtain the desired 30

respondents, and 100 families were mailed the alternative survey.


Apartment Dwellers

Using a list of apartment developments in the Gainesville metropoli-

tan area, 10 apartment complexes were used to obtain the sample of

apartment dwellers. Using a random numbers table, the sample for both

survey methods was obtained from a list of utility customers obtained









from the Regional Utilities Board. A pretest of the apartment dwellers

indicated that a response rate of approximately 20 percent for personal

interviews and 30 percent for mail-out questionnaires could be expected;

therefore, 150 families were chosen randomly to obtain 30 respondents

for the personal interviews and 100 families were chosen for the

alternative survey.

The families selected for the personal interviews were mailed a

letter describing the study (see Appendix A). They were contacted two

weeks prior to the desired time for interviews to set up appointments.

Since studies have shown the response rate is not significant when

the sample is drawn from a homogenous population [Tull and Hawkins, 1976],

the low response rate in some consumer groups is assumed not to have

caused bias in the responses.


Instruments

Personal Interview Survey with Prices

Using various preference categories from previous studies, an

instrument with 24 responses was developed by the writer (See Appendix B).

The responses given by the respondents were recorded on the survey form.

Each question required the respondents to make a decision based on price,

space, space trade-off or a combination of the three items.

A maximum price of $38,500 was assumed for the preferences indicated.

Responses were recorded and if the sum of all feature options exceeded

$38,500, respondents were instructed to reduce choices according to price

and utility until the maximum allowable price was obtained.









Mail-Out Questionnaires

The alternative survey instrument was mailed to the homeowners and

apartment dwellers. The questions contained in the mail-out question-

naires sought the same responses contained in the surveys used in the

personal interviews (see Appendix C).

The mail-out questionnaire instructed the respondents to assume a

maximum price of $38,500, and to use their intuitive knowledge of housing

costs to make various choices. This instrument did not contain the cost

information provided the respondents in the personal interviews. One

additional question on this instrument required the respondents to rank

several feature options, if it was necessary to give up some choices in

order to maintain the maximum price of $38,500.


Demographic Information

The last section of both survey instruments contained questions that

were designed to identify different market segments (see Appendix C).

The demographic characteristics used were identified by market researchers

as potentially useful in segmenting markets of consumer durables

[Tull and Hawkins, 1976].

The last question of the homeowners' questionnaire sought to

identify specific areas of owners displeasure with their present homes.

Although this information was not pertinent to this study, local

builders were interested in knowing of homeowners' dissatisfaction with

their newly purchased home.


Collection of Data


The collection of data was accomplished using the two survey forms

developed by the writer. The homeowners and apartment dwellers were









surveyed utilizing two different techniques- mail-out questionnaires

and personal interviews.

As previously stated, homeowners and apartment dwellers selected to

participate in the survey using price information were contacted using a

letter describing the study. These two groups were contacted and

appointments were set up to administer the survey using the personal

interview technique. Using a graphic display to assist the respondents

in making their choice, the interviewer recorded each response on a

prepared survey form (see Appendix B). To minimize interviewer bias,

the survey forms contained the questions to be asked, and the inter-

viewers were instructed not to give any additional information not con-

tained in survey form narrative or information contained in the graphic

display kit.

The other group of homeowners and apartment dwellers were mailed

the questionnaire accompanied by a cover letter explaining the study.

The respondents were instructed to answer the questions based on the

instructions contained in the questionnaire. The writer allowed three

weeks for respondents to return the questionnaires. During this period

of time a sufficient number of questionnaires were returned so no

follow-up was used. In both survey techniques the respondents were

insured of anonymity. When all of the survey forms were completed for

both groups using both survey instruments, the data were transferred to

punch cards for computer analysis.


Analysis of Data

The first step in the analysis of data was to create frequency

tables for each consumer group by survey type. On the basis of the









indicated choice for feature options, the two groups (homeowners and

apartments) were combined to test for significant differences in prefer-

ences yielded by the two different survey instruments.

Cross tabulation for selected demographic characteristics was also

performed. This was accomplished by combining the two groups of

respondents and comparing the results by survey instrument type.

It was determined that because of the numerical counting of the

responses or frequency of occurrence, Chi-square tests would be

appropriate to test for significance differences in preferences

yielded by the two survey techniques. Using a Statistical Analysis

System program, contingency tables were constructed that calculated

Chi-square values and the level of significance.

Various cross-tabulations were performed using demographic

characteristics. Cross-tabulations of consumer type and survey form

were also performed.

Using the total price of the respondents choices obtained from

both survey methods, a test of differences between the sample means

was made. This test was made as a secondary check of the results

obtained in the Chi-square tests.

Additional frequency tables were constructed to identify the order

of trade-off when respondents had to reduce costs because of exceeding

the maximum price ($38,500).














CHAPTER IV

PRESENTATION AND ANALYSIS OF DATA


The primary objective of this study is to determine whether the

knowledge of price affects consumers' preferences of optional housing

features. Responses obtained from homeowners and apartment dwellers

were used as a proxy for consumers' preferences related to housing

feature options. The specific hypothesis tested is

There is no significant difference in the preferences of home
buyers who are aware of the price of each optional feature and
home buyers who are not aware of the price.

The feature options used in the study were identified in the

literature as those that consumers considered to be important (see

Chapter II). An attempt was made to limit the number of feature options

available so that the options offered the respondents were cost feasible

within the upper limit price constraint of the study.

Families living in single-family homes and apartments were surveyed

using an instrument showing the price of each choice; additionally, the

graphic display kit aided the respondents to inspect each choice

visually. The other group was surveyed using mail-out questionnaires.

Although this instrument sought the same responses, this group had no

knowledge of price directly from the survey form. As in previous

studies, they were asked to use a given price constraint and indicate

their choices subject to this price (see Chapter II).









Twenty-four areas of feature options were included in the study.

Preferences for each option were based upon respondents' indications of

their choices for each area.

Response 1. Importance of space allocation in the home.

Response 2. Number of bedrooms.

Response 3. Number of bedrooms used for purposes other than
sleeping.

Response 4. Master bedroom size.

Response 5. Secondary bedroom size.

Response 6. Number of baths.

Response 7. Type of shower/tub and enclosure material.

Response 8. Kitchen size and cabinet quality.

Response 9. Kitchen options.

Response 10. Dining space alternatives.

Response 11. Living area size.

Response 12. Kitchen/dining and living area layout alternatives.

Response 13. Space trade-off for increased dining area.

Response 14. Exterior construction type.

Response 15. Lot size.

Response 16. Parking facility alternatives.

Response 17. Floor covering.

Response 18. Interior finish options.

Response 19. Heating and cooling options.

Response 20. Storage options.

Response 21. Exterior options.

Response 22. Insulation options.

Response 23. Energy saving options.

Response 24. Landscaping packages.









Data from 118 respondents provided the input for the Chi-square

tests.

The responses of 61 respondents using price in their decision-

making were compared with the responses of 57 respondents using only

their intuitive knowledge of housing costs for significant differences.

A 0.05 level of significance was used in all Chi-square tests. In

cases where cell size of fewer than five was achieved, the results were

considered to be inconclusive [Richmond, 1969].

The contingency tables presented indicate that the preference of

the respondents from both groups are compared based on different

methodologies. Apartment dwellers and homeowners are combined for

this test.


Response 1. Importance of space allocation in the home.


A B TOTAL

C 20 30 55

N 8 41 49

TOTAL 28 76 104

C = Cost survey form

N = No cost survey form (Mail-out questionnaire)

Chi-square = 5.28 D.F. = 1

Prob. = 0.0215

Although cost was not a factor in this decision, the researcher

must reject the hypothesis that the preferences of both groups are the

same. The respondents were asked to indicate a preference for more

space to be allocated in the bedroom/bath area (A) or in the









kitchen/dining and living area (B). At the 0.05 level, this hypothesis

can be rejected.


Response 2. Number of bedrooms.


A B C TOTAL

C 4 45 6 55

N 6 41 4 51

TOTAL 10 86 10 106

Chi-square = 0.836 D.F. = 2

Prob. = 0.6583


The respondents were given three choices for the number of bed-

rooms, 2 (A), 3 (B), and 4 (C). Although the hypothesis cannot be

rejected at the 0.05 level, the results are inconclusive because of the

two cells containing fewer than five observations.


Response 3. Number of bedrooms used for purposes other
than sleeping.


N Y TOTAL

C 21 34 55

N 8 43 51

TOTAL 29 77 106

Chi-square = 6.738 D.F. = 2

Prob. = 0.0094


The hypothesis can be rejected at the 0.05 level that the two

groups tested have the same preference regarding bedroom placement.

More importantly, almost 75 percent of the respondents indicated that

at least one bedroom (Y) would be used for purposes other than sleeping,









while 25 percent (X) indicated that all bedrooms would be used for

sleeping. Builders have noted that when "bedrooms" are used for other

purposes, it allows more flexibility in design [Consumer Attitude

Study, 1977].


Response 4. Master bedroom size.


L S TOTAL

C 49 6 55

N 49 2 51

TOTAL 98 8 106

Chi-square = 1.852 D.F. = 1

Prob. = .1736


The hypothesis cannot be rejected at the 0.05 level that the two

groups have different preferences regarding master bedroom size;

however, caution should be exercised in determining precisely the level

of confidence because of the one small cell size. It is important to

note that over 90 percent of both groups prefer the large master

bedroom (at least 12 feet by 14 feet).


Response 5. Secondary bedroom size.


A Chi-square test was not possible for testing differences in

secondary bedroom size. Respondents were given five bedroom sizes to

choose from; however, most respondents chose two different size bed-

rooms. The five different size bedrooms produced 14 different combina-

tions, with many cells having zero observations. Based on visual

inspection of frequency tables, very little difference exists between








the two groups; additionally, consumers seem to prefer a combination of

a small or medium size bedroom and a larger bedroom.


Response 6. Number of baths.


B D TOTAL

C 37 18 55

N 28 23 51

TOTAL 65 41 106

Chi-square = 1.707 D.F. = 1

Prob. = 0.1913


To insure a valid test, choices of bath(s) had to be combined.

Choice A (one bath) and Choice B (one and one-half bath) were combined

into one choice (B) and Choice C (one and three-quarters baths) and

Choice D (two baths) were combined into the other choice (D). The

hypothesis that there is no difference in consumer choice between the

two consumer groups cannot be rejected at the 0.05 level.


Response 7. Type of shower/tub and enclosure material.


F T TOTAL

C 36 19 55

N 14 37 51

TOTAL 50 56 106

Chi-squared = 15.337 D.F. = 1

Prob. = 0.0001

The respondents were asked to indicate preferences regarding tub

type and enclosure material. The hypothesis can be rejected at the









0.05 level that there is no difference in the preferences of the two

groups surveyed. The group (C) that was aware that there was a substan-

tially lower price of the fiberglass molded tub/enclosures, chose the

fiberglass units almost 2:1 over the cast iron tub with ceramic tile

walls. The group (N), not aware of the price differential, chose the

cast iron tub with ceramic tile 2:1 over the fiberglass.


Response 8a. Kitchen size.


C

N

TOTAL


L S TOTAL

42 13 55

39 12 51

81 25 106

Chi-square = 0.000 D.F. = 1

Prob. = .9897


The Chi-square value of 0.00 indicates the difference between the

choices of the two groups for large (L) and small (S) kitchensis not

calculable; therefore, the hypothesis that there is no difference in

the groups' choice cannot be rejected at the 0.05 level.


Response 8b. Cabinet quality.


C

N

TOTAL


69 37

Chi-square = 7.778

Prob. = 0.0487


TOTAL

55

51

106

D.F. = 1









Respondents aware of the substantial price difference (C) were

split in the decision regarding the type of cabinet they preferred;

however, the other group of respondents (N) selected wood almost 4:1

over the plastic laminated cabinets. The hypothesis that there is no

difference in preferences related to cabinet quality can be rejected

at the 0.05 level of confidence.


Response 9a. Kitchen option dishwasher.


O 1 TOTAL

C 4 51 55

N 9 42 51

TOTAL 13 93 106

Chi-square 2.647 D.F. = 1

Prob. = 0.1038


Response 9b. Kitchen option garbage disposal.


O 1 TOTAL

C 14 41 55

N 17 34 51

TOTAL 31 75 106

Chi-square = 0.794 D.F. = 1

Prob. = 0.3729









Response 9c. Kitchen option self-cleaning oven.


O 1 TOTAL

C 31 24 55

N 20 31 51

TOTAL 51 55 106

Chi-square = 3.117 D.F. = 1

Prob. = 0.0775


Response 9d. Kitchen option microwave oven.


O 1 TOTAL

C 51 4 55

N 42 9 51

TOTAL 93 13 106

Chi-square = 2.647 D.F. = 1

Prob. = 0.1038


Response 9e. Kitchen option countertop backsplash.


0 1 TOTAL

C 26 29 55

N 20 31 51

TOTAL 46 60 106

Chi-square = 0.699 D.F. = 1

Prob. = 0.4030


Based on the Chi-square tests performed, there is no significant

difference in the preferences of the two consumer groups related to

the five kitchen options; therefore, the hypothesis cannot be rejected









at the 0.05 level of confidence. More than 80 percent of the respon-

dents indicated that they wanted the dishwasher and garbage disposal;

likewise, more than 85 percent did not want the microwave oven. The

microwave oven had been identified by other studies (Chapter II) as

a new appliance being demanded by new home purchasers.


Response 10. Dining area.


L N TOTAL

C 36 19 55

N 26 25 51

TOTAL 62 44 106

Chi-square = 2.283 D.F. = 1

Prob. = 0.1308


No significant difference in preferences related to dining area

was found between groups. Respondents were asked to choose between an

area included in kitchen for dining and a separate dining area. While

62 respondents preferred a separate (L) area and 44 respondents

preferred an "eat-in area," no conclusive evidence can be inferred

from the data; therefore, the hypothesis cannot be rejected at the

stated level of confidence that there is no difference in preferences

of the two groups related to dining area.









Response 11. Living area.


A B C TOTAL

C 0 15 40 55

N 7 13 31 51

TOTAL 7 28 71 106

Chi-square = 0.062 D.F. = 1

Prob. = 0.8030


In addition to the test that shows no significant difference in

preferences of the two groups, the data show that almost 70 percent of

the respondents chose the large living area. This result indicates

that home buyers will give up space in other areas to have a large

"family room."


Response 12. Kitchen/dining and living area layout alternatives.


A B C TOTAL

C 16 10 29 55

N 10 12 29 51

TOTAL 26 22 58 106

Chi-square = 1.418 D.F. = 2

Prob. = 0.4923


After choosing dining area, kitchen size, and living/family area,

the respondents were asked to choose between three typical layouts

that combine these three areas in the kitchen/dining and family area.

The Chi-square test indicated no significant difference in the prefer-

ences of the two groups when asked to choose between the open plan (A),

country kitchen (B), and zoned plan (C).








Response 13. Space trade-off for increased dining area.


N Y TOTAL

C 45 10 55

N 34 17 51

TOTAL 79 27 106

Chi-square = 3.200 D.F. = 1

Prob. = 0.0736


When asked whether they would be willing to sacrifice space in

other areas of the home for increased dining area, over 65 percent

indicated they would not be willing (N). No significant difference

between the two groups' preferences was found.


Response 14. Exterior construction type.


B N TOTAL

C 20 35 55

N 15 36 51

TOTAL 35 71 106

Chi-square = 0.578 D.F. = 1

Prob. = 0.4470


Respondents were given three choices for materials used on

exterior construction- brick veneer, wood frame (wood siding), and

concrete block. For purposes of analysis the two groups surveyed were

tested for a difference in preference for brick (B) or other type of

construction, wood frame and concrete block (N). No significant dif-

ference in material type for exterior construction was found between

the two groups.









Response 15. Lot size.


A B C TOTAL

C 10 22 21 53

N 5 8 38 51

TOTAL 15 30 59 104

Chi-square = 13.865 D.F. = 2

Prob. = 0.0010


Over 75 percent of the respondents from the mail-out questionnaire

indicated a preference for the large lot (C). These home buyers were

not aware of the substantial price differences among the three lots.

As indicated from the test results, a significant difference in prefer-

ences was found between the two groups; therefore, the hypothesis that

no difference in preferences exists can be rejected at the 0.05 level.


Response 16. Parking facility alternatives.


G N TOTAL

C 47 8 55

N 40 11 51

TOTAL 87 19 106

Chi-square = 0.887 D.F. = 1

Prob. = 0.3462


Respondents were given four different automobile parking alterna-

tives. To increase the reliability of the Chi-square test, some of the

choices had to be grouped so that the cell size was five or greater.

The test involved the choice of a garage or a carport or parking slab.

From the contingency table, it is evident that no significant difference









was found between the two groups; therefore, the hypothesis that there

is no significant difference in preferences related to parking alterna-

tives cannot be rejected at the 0.05 level of confidence.


Response 17. Floor covering quality.


0 1 TOTAL

C 50 6 56

N 18 33 51

TOTAL 68 39 107

Chi-square = 28.416 D.F. = 1

Prob. = .0105


The respondents were asked to indicate their preferences regarding

floor covering materials in all areas of the home. No observable dif-

ferences in the type of floor covering desired in the various rooms of

the home were found. The typical responses were carpet in all bedrooms,

living area, dining area, hall, tile or ceramic tile in the baths, and

vinyl tile or linoleum in the kitchen.

A test was made to determine whether there was a significant difference

in the preferences for floor covering quality. Group N indicated almost

2:1 that they preferred the higher quality floor coverings. They wanted

the higher quality carpeting in the areas where carpet was preferred,

ceramic tile in the bath, and vinyl linoleum in the kitchen. The

hypothesis that the two groups have similar preferences regarding floor

covering quality can be rejected at the 0.05 level.









Response 18a. Interior finish option paneling.


0 1 TOTAL

31 24 55

34 17 51

65 41 106

Chi-square = 1.184 D.F. = 1

Prob. = 0.2765


Response 18b. Interior finish option beamed cathedral ceiling.


28

26

54

Chi-square


27

25

52

= 0.000


TOTAL

55

51

106

D.F. = 1


Prob. = 0.9941


Response 18c. Interior finish option bay window.


0 1 TOTAL

C 29 26 55

N 26 25 51

TOTAL 55 51 106

Chi-square = 0.032 D.F. = 1

Prob. = 0.8573


C

N

TOTAL


C

N

TOTAL








Response 18d. Interior finish option wallpaper package for
baths and kitchen.


O 1 TOTAL

C 22 33 55

N 12 39 51

TOTAL 34 72 106

Chi-square = 3.295 D.F. = 1

Prob. = 0.0695


Response 18e. Interior finish option tile or simulated marble
window sills.


O 1 TOTAL

C 41 14 55

N 27 24 51

TOTAL 68 38 106

Chi-square = 5.271 D.F. = 1

Prob. = 0.0205


Response 18a e. Interior options.


Five interior options were offered to the respondents in both con-

sumer groups. The preferences of the two groups were significantly dif-

ferent for one option, tile or simulated marble window sills. The

hypothesis can be rejected at the 0.05 level that no significant differ-

ence exists between the two groups related to the option tile or

simulated marble window sills.









Response 19a. Heating and cooling option central heat only.


0 1 TOTAL

C 53 2 55

N 51 0 51

TOTAL 104 2 106

Chi-square = 1.890 D.F. = 1

Prob. = 0.1692


Response 19b. Heating and cooling option central air and
heating.


O 1 TOTAL

C 4 51 55

N 1 50 51

TOTAL 5 101 106

Chi-square = 1.661 D.F. = 1

Prob. = 0.1974


Response 19c. Heating and cooling option exhaust fan.


0 1 TOTAL

C 43 12 55

N 36 27 51

TOTAL 79 37 106

Chi-square = 0.804 D.F. = 1

Prob. = 0.3700









Response 19d. Heating and cooling option wood stove.


0 1 TOTAL

C 46 7 55

N 50 1 51

TOTAL 96 8 106

Chi-square = 4.396 D.F. = 1

Prob. = 0.0360


Response 19e. Heating and cooling option prefab fireplace.


C

N

TOTAL


0 1 TOTAL

35 20 55

48 6 54

83 26 109

Chi-square = 14.471 D.F. = 1

Prob. = 0.0001


Response 19f. Heating and cooling option masonry fireplace.


0 1 TOTAL

C 46 9 55


N

TOTAL


21 30

67 39

Chi-square = 20.514

Prob. = 0.0001


51

106

D.F. = 1









Response 19a f. Heating and cooling options.


As expected, respondents from both groups indicated that they

wanted central air and heating. More than 95 percent of the respondents

indicated their preference for this option (Response 19b).

Responses 19e (prefab fireplace) and 19f (masonry fireplace)

yielded conclusive evidence that depending on presence or absence of

price information, respondents have almost opposing preference.

Respondents from Group C (knowledgeable of price), who indicated they

wanted a fireplace, chose the significantly less expensive prefab

fireplace almost 3:1 over the masonry fireplace. Respondents from the

mail-out questionnaire (no knowledge of price) chose the masonry fire-

place over the prefab fireplace 5:1.

The hypothesis that there is no difference in the preferences of the

two consumer groups can be rejected at the 0.05 level.


Response 20a. Storage option walk-in closet.


0 1 TOTAL

C 41 14 55

N 20 31 51

TOTAL 61 45 106

Chi-square = 13.520 D.F. = 1

Prob. = 0.0002









Response 20b. Storage option additional closet (not walk-in).


O 1 TOTAL

C 45 10 55

N 39 12 51

TOTAL 84 22 106

Chi-square = 0.460 D.F. = 1

Prob. = 0.4976


Response 20c. Storage option additional shelving (garage).


O 1 TOTAL

C 36 19 55

N 34 17 51

TOTAL 70 36 106

Chi-square = 0.017 D.F. = 1

Prob. = 0.8953


Response 20a c. Storage options.


Three storage options were available to the respondents (shown in

Responses 20a, b, and c). One option, the walk-in closet, exhibited a

significant difference in group preferences. Over 60 percent of Group

N (no knowledge of its cost) indicated a preference for the walk-in

closet, while fewer than 25 percent of Group C (knowledge of price and

space trade-off) chose this option. The hypothesis that there is no

difference in preferences between the groups can be rejected at the

0.05 level.









Response 20 la. Space trade-off for walk-in closet master
bedroom.


1 TOTAL

4 55


31 20

82 24

Chi-square = 15.416

Prob. = 0.0001


51

106

D.F. = 1


Response 20 lb.


Space trade-off
ary bedrooms.


for walk-in closet second-


0 1

47 8

36 15

83 23

Chi-square = 3.442

Prob. = 0.0636


Response 20 Ic.


Space
area.


52


50 3

102 6

Chi-square = 0.890

Prob. = 0.3456


trade-off for walk-in closet living



1 TOTAL

3 55


51

106

D.F. = 1


C

N

TOTAL


C

N

TOTAL


TOTAL

55

51

106

D.F. = 1


C

N

TOTAL









Response 20 Id. Space trade-off for walk-in closet kitchen/
dining area.


0 1 TOTAL

C 53 2 55

N 49 2 51

TOTAL 102 4 106

Chi-square = 0.006 D.F. = 1

Prob. = 0.9386


Responses 20 la d. Space trade-off for walk-in closet.


If the respondents indicated a preference for a walk-in closet,

a space trade-off was necessary. They were to indicate the area of the

home in which they would be willing to give up space to accommodate the

walk-in closets. The choices are shown in the preceding contingency

tables. If they were willing to give up space in an area, the observa-

tion was counted as a positive response (1); negative response was

listed as "O". The hypothesis that there is no difference in preference

for space trade-off cannot be rejected at the 0.05 level.


Response 21a. Exterior option patio with sliding glass door.


0 1 TOTAL

C 11 44 55

N 15 36 51

TOTAL 26 80 106

Chi-square = 1.266 D.F. = 1

Prob. = 0.2605









Response 21b. Exterior option patio cover.


29 26

40 11

69 37

Chi-square = 7.695

Prob. = 0.05


TOTAL

55

51

106

D.F. = 1


Response 21c. Exterior option patio screen.


0 1

C 40 15

N 31 20

TOTAL 71 35

Chi-square = 1.707

Prob. = 0.1914


TOTAL

55

51

106

D.F. = 1


Response 21d. Exterior option gutters.


0 1

36 19

25 26

61 45

Chi-square = 2.926

Prob. = 0.0872


TOTAL

55

51

106

D.F. = 1


C

N

TOTAL


C

N

TOTAL








Response 21e. Exterior option higher quality shingles.


O 1 TOTAL

C 49 6 55

N 36 15 51

TOTAL 85 21 106

Chi-square = 5.703 D.F. = 1

Prob. = 0.0169


Response 21f. Exterior option aluminum soffits.


O 1 TOTAL

C 34 21 55

N 44 7 51

TOTAL 78 28 106

Chi-square = 8.143 D.F. = 1

Prob. = 0.0043


Response 21a f. Exterior options.


Six exterior options were available to the respondents. The

Chi-square test revealed that the preferences for three of the options

were significantly different between the two groups surveyed.

Two exterior options produced unexpected significant differences.

Although they were knowledgeable of the price of the options, Group C

had significantly different preferences for patio covers and aluminum

soffits. It is the opinion of the author that this significant

difference can be explained by a lack of information on the part of

Group N. Several respondents from Group N indicated questions

regarding these two options. Group C, with the aid of the graphic









display kit, was able to see potential benefits derived from these two

options.

The preference for higher quality shingles was also significantly

different for the two groups. Respondents in Group C indicated that

they were not willing to incur additional cost for shingles that would

last five to ten years longer because they did not anticipate living in

the home beyond the expected life of the standard shingles.


Response 22. Insulation options.


A B TOTAL

C 15 40 55

N 5 46 51

TOTAL 20 86 106

Chi-square = 6.564 D.F. = 1

Prob. = 0.0104


Two insulation alternatives were available to the respondents.

Alternative A was standard and alternative B was additional insulation

to create a more energy efficient home. Fewer than 75 percent of Group

C preferred the additional insulation, while over 90 percent of Group

N preferred the additional insulation; therefore, the hypothesis that

there is no significant difference between the preferences of the two

groups regarding insulation alternatives can be rejected at the 0.05

level.








Response 23a. Energy option turbine vents.


0 1 TOTAL

32 23 55

30 21 51

62 44 106

Chi-square = 0.004 D.F. = 1

Prob. = 0.9466


Response 23b. Energy option ridge vent.


C 43 12

N 44 7

TOTAL 87 19

Chi-square = 1.178

Prob. = 0.2778


Response 23c. Energy option -
fan.


0 1


C

N

TOTAL


49 6

30 21

79 27

Chi-square = 12.77(

Prob. = 0.0004


TOTAL

55

51

106

D.F. =1




thermostatically controlled attic



TOTAL

55

51

106

0 D.F. = 1


C

N

TOTAL









Response 23d. Energy option insulated windows.


C

N

TOTAL


0 1 TOTAL

45 10 55

20 31 51

65 41 106

Chi-square = 20.249 D.F. = 1

Prob. = 0.0001


Response 23e. Energy option hot water heater timer.


C

N

TOTAL


0 1 TOTAL

41 14 55

36 15 51

77 29 106

Chi-square = 0.209 D.F. = 1

Prob. = 0.6479


Response 23f. Energy option solar water heater.


0 1

C 51 5

N 18 33

TOTAL 69 38

Chi-square = 38.416

Prob. = 0.0001


TOTAL

56

51

107

D.F. = 1


Response 23a f. Energy options.


The literature (Chapter II) indicated that consumers are becoming

more aware of potential energy savings through the use of various new









products. Six energy options were included in the survey (see preced-

ing contingency tables). All of the energy options are those that would

be cost feasible for a home in the price range of the study. The price

of the options ranged from $150 to $1,500.

Preferences for three of the energy options produced significant

differences between the two groups surveyed. Respondents in Group N

indicated a preference for a thermostatically controlled attic fan and

insulated windows 3:1 over the respondents in Group C.

Respondents in Group N preferred the solar water heater 6:1 over

the respondents in Group C. This option, costing $1,500, did not appeal

to the respondents in Group C because they did not think that the price

of the solar water heater could be recovered in energy savings during

the period of time they would be occupying the home.


Response 24. Landscaping packages.


0 B X TOTAL

C 2 17 36 53

N 1 20 30 50

TOTAL 3 37 66 103

Chi-square = 0.702 D.F. = 1

Prob. = 0.4021


In the survey, the subjects were given three choices for land-

scaping. Each choice had different items provided by the builder at a

specified cost. To insure that a valid test would be obtained, two

landscape packages were combined because of small cell size (less than

five); therefore, for purposes of analysis, the two groups were tested

for differences based on the basic package (B) and an extra package (X).









The hypothesis that there is no difference in the preferences of the

two groups cannot be rejected at the 0.05 level.


Summary of Individual Differences


The individual tests for preference differences between groups

indicate significant differences in several areas. The following table

summarizes these differences.









TABLE 1

SUMMARY OF INDIVIDUAL PREFERENCE DIFFERENCES


Area of difference Significance level


Importance of space allocation in the home 0.0215

Use of bedrooms for purposes other than sleeping 0.0094

Type of shower/tub and enclosure material 0.0001

Kitchen cabinet quality 0.0487

Lot size 0.0010

Floor covering quality 0.0105

Interior option tile or simulated marble window sills 0.0205

Heating and cooling options

Wood stove 0.0360

Prefab fireplace 0.0001

Masonry fireplace 0.0001

Storage option walk-in closet 0.0002

Exterior options

Patio cover 0.0500

Higher quality shingles 0.0169

Aluminum soffits 0.0043

Insulation options 0.0104

Energy options

Thermostatically controlled attic fan 0.0004

Insulated windows 0.0001

Solar water heater 0.0001








Although these individual differences were found to exist, further

tests were made to determine whether these differences resulted in dif-

ferences in the price of homes.


TABLE 2
TOTAL PRICE OF HOME NO KNOWLEDGE OF PRICE OF OPTIONAL FEATURES


Homeowners


Apartment Dwellers


$39,655
$44,790
$38,813
$43,640
$42,440
$40,270
$38,480
$40,750
$43,170
$42,225
$47,610
$43,215
$42,330
$41,550
$42,190
$43,395
$49,715
$41,165
$44,545
$41,690
$45,065
$44,275
$44,465
$43,580
$42,805
$42,125
$38,475
$43,100


43,190
45,020
46,885
40,720
41,885
48,220
45,160
47,080
38,600
44,920
48,015
38,700
43,195
46,010
47,160
38,800
41,700
41,515
43,380
40,195
43,190
44,760
46,265
41,850
43,980
44,805
47,510










TOTAL PRICE OF HOME -


Homeowners


$39,755
$37,165
$38,000
$39,910
$38,790
$40,255
$39,990
$40,710
$39,668
$39,680
$39,570
$39,795
$38,480
$39,330
$39,280
$39,350
$40,270
$41,320
$39,325
$38,970
$39,590
$40,290
$40,110
$41,190
$39,790
$38,470
$39,620
$40,120
$38,690
$41,910
$39,520


TABLE 3

KNOWLEDGE OF PRICE OF OPTIONAL FEATURES


Apartment Dwellers


40,807
38,050
42,654
39,396
39,269
38,648
38,408
38,778
40,347
38,394
38,838
40,966
38,326
39,394
38,247
38,667
38,345
38,652
42,129
40,601
39,125
38,343
38,731
38,595
39,308
38,787
38,455
38,979
39,173
38,338








Test of Differences Between Sample Means

In the preceding section, the results of Chi-square tests on

individual choices of respondents in both groups surveyed are reported.

The tests revealed significant differences of preferences of 16 responses

that were cost related. Other preference differences related to room

size and layout arrangement were also found to exist between the two

groups of respondents.

To determine whether the differences found in the individual

preferences between the groups produced differences in the price of

the home desired, a test of differences between the sample means was

made. The gross total price (no price trade-off) of the respondents

of each group was determined by summing the cost of all preferences.

Using the data presented in Tables 2 and 3, the mean and standard

deviation were calculated. The following data were used to test the

hypothesis that the mean of the two samples were the same.


H : 2 = 0


Ha: 2 > 0


Group N Group C

n1 = 57 n2 = 61

Xl = $43,240 2 = $39,470

s = $2,744 2 = $1,026


ni = number of observations in sample

X. = sample mean

si = standard error of the mean








The observed difference in the sample means was $3,770. To test

for significant differences between sample means, the standard error

of the difference between the two means was calculated.



F12 2
sd sx1 sx2 n n2



S (2,744)2 (1026)2


= /149,35


sd = 386.46


The calculated Z-value was computed and compared with the critical

value of 1.96 (0.05 level).


1z
Zs
sd


S$43,240 $39,470
386.46

3770
386.46

Z = 9.755

The calculated Z-value of 9.755 exceeds the critical value of
1.96; therefore, the null hypothesis can be rejected at the 0.05

level. By rejecting the hypothesis that the difference of the sample








means is equal to zero, it is implied that the difference in the means

is attributed to factors other than chance.

Based on the above test, it can be concluded that the respondents

of each group had significantly different preferences that resulted in

a differently priced home; furthermore, Group N (no knowledge of the

price of each preference) had the highest mean priced home. This

would indicate that knowledge of the price (Group C) allows respondents

to choose feature options most suited to their budgetary constraints.


Test of Differences Between Homeowners'
and Apartment Dwellers' Preferences

Related to the primary objective of this study, this section seeks

to test for differences in preferences exhibited by apartment dwellers

and homeowners. Builders realize that persons who have not owned

homes previously (first-time buyers) may have needs and desires differ-

ent from home buyers who have previously owned a home [Goodkin, 1974].

The first-time buyers are usually young singles or marrieds, while the

homeowners are typically young marrieds with children or established

families [Consumer Attitude Survey, 1977]. With such a diverse market,

builders must be able to divide these groups into separate markets

(market segmentation) and identify preference differences of the groups.

Using the data collected in the personal interviews, Chi-square

tests were made for each choice to determine individual preference

differences between homeowners and apartment dwellers seeking to pur-

chase their first homes. The homeowner sample contained 31 observations,

and the apartment dweller sample contained 30 observations. Significant

differences were found in nine preference categories.








Response 1. Importance of space allocation in the home.


A B TOTAL

A 5 23 28

H 17 12 29

TOTAL 22 35 57

Chi-square = 13.133 D.F. = 1

Prob. = 0.0003

Significant differences of preferences were found to exist in

space allocation within the home. Over 80 percent of the apartment

dwellers (A) preferred that more space be allocated to the kitchen/

dining and living areas of the home. This is consistent with the

findings of other studies. One study determined that the design

implications for first-time buyers were for living area/dining room

combination emphasizing usable space [Consumer Attitude Study, 1977].

The homeowners indicated a different preference, with 60 percent

preferring more space in the bedroom/bath area. This finding is also

consistent with other studies. This group, in previous studies, was

found to place as much emphasis on bedroom size as on the number of

bedrooms [Consumer Attitude Study, 1977].









Response 2. Number of bedrooms.


A B C TOTAL

A 4 22 0 26

H 0 23 6 29

TOTAL 4 45 6 55

Chi-square = 9.888 D.F. = 2

Prob. = 0.0071


The test indicates that a significant difference exists between

the groups related to the number of bedrooms preferred; however, with

one cell containing fewer than five observations, it is not possible to

determine the precise level of significance. Design information can be

gleaned from the data, however. In both samples more than 80 percent

of the respondents indicated three bedrooms (B) as their preference.

The significant difference was calculated on the basis of four apart-

ment dwellers' (A) preferring two bedrooms and six homeowners' prefer-

ring four bedrooms. The findings are consistent, however, in that

apartment dwellers who did not prefer three bedrooms want two bedrooms,

and homeowners who did not want three bedrooms preferred four.


Response 6. Number of baths.


B D TOTAL

A 22 6 28

H 15 14 29

TOTAL 37 20 57

Chi-square = 6.736 D.F. = 1

Prob. = 0.0094








The respondents had a choice of four bath options: one, one and

one-half, one and three-quarters, and two baths. To insure a valid

test, the first two choices were combined into one choice (B) and

choices three and four were combined to form a single choice (D).

Eighty percent of the apartment dwellers (A) indicated a preference

for one or one and one-half baths. Fifty percent of the homeowners

chose option (D).

Several homeowners responding to this question indicated that they

had one bath or one and one-half baths presently and when they purchased

the home thought this would serve their needs; however, after living in

their home for a time, they felt they needed additional bath space.

The data presented provided evidence that a significant difference

exists between the two groups. The hypothesis that the groups surveyed

have similar preferences for baths can be rejected at the 0.05 level.


Response 9b. Kitchen option garbage disposal.


0 1 TOTAL

A 12 14 26

H 5 27 32

TOTAL 17 41 58

Chi-square = 11.134 D.F. = 1

Prob. = 0.0005









Response 9e. Kitchen option counter top backsplash.


0 1 TOTAL

A 6 20 26

H 20 9 29

TOTAL 26 29 55

Chi-square = 11.582 D.F. = 1

Prob. = 0.0007


The respondents were asked to indicate their preferences for five

kitchen options. In the data presented, significant differences for

two of the kitchen options were found.

The difference in the preferences for the counter top backsplash

may be attributed to experience of the homeowners. Several homeowners,

without the backsplash in their present homes, indicated their prefer-

ence to include it in their next home for obvious reasons; the back-

splash reduces maintenance in the work area of the kitchen.


Response 12. Kitchen/dining and living area layout alternatives.


A B C TOTAL

A 6 5 19 30

H 10 8 10 28

TOTAL 16 13 29 58

Chi-square = 10.059 D.F. = 2

Prob. = 0.0065

A = Open plan

B = Country kitchen

C = Zoned plan








Three layout alternatives for the kitchen/dining and living area

were offered to the respondents. No distinct preference was found for

each alternative in either group; however, significant differences were

found when the two groups were compared.

The hypothesis that the two groups have similar preferences for

layout arrangement can be rejected at the 0.05 level. The data do not

present any conclusive evidence for design information. It would appear

from observation that choice (C), with 50 percent, is the most popular

layout.


Response 14. Exterior construction type.


B N TOTAL

A 17 9 26

H 5 26 31

TOTAL 22 35 57

Chi-square = 17.947 D.F. = 1

Prob. = 0.0001

B = Brick veneer exterior

N = Non-brick exterior (wood siding or concrete
block)

The preferences for exterior construction were reduced to brick

veneer exterior or non-brick exterior. The preferences indicated by

the two groups were almost opposite; therefore, the hypothesis is

rejected that the two groups have similar preferences for exterior

construction at the 0.05 level.

Many of the homeowners surveyed indicated that because of the

substantial cost difference, they had chosen a non-brick exterior for









their present home. Many expressed satisfaction with the wood siding

that was used in the home construction.


Response 18c. Interior option bay window.


A

H

TOTAL


7 22

29 29

Chi-square = 13.173

Prob. = 0.0003


TOTAL

29

29

58

D.F. =1


Response 18d. Interior option -
and kitchen.


A

H

TOTAL


wallpaper package for baths


0 1 TOTAL

15 11 26

7 22 29

22 33 55

Chi-square = 6.431 D.F. = 1

Prob. = 0.0112


Five choices were available to the respondents in the interior

options area. Significant differences in preferences between the

groups were found for two options presented in the preceding contin-

gency tables.

Consistent with the findings of other studies, homeowners indicated

a preference for decorative options [Consumer Attitude Study, 1977].

Builders attempting to respond to the preferences of families changing

homes can add decorative, eye-appealing options with little additional

cost.









Response 19e. Heating and cooling option prefab fireplace.


0 1 TOTAL

A 21 5 26

H 14 15 29

TOTAL 35 20 55

Chi-square = 6.255 D.F. = 1

Prob. = 0.0124


One-half of the homeowners indicated a preference for the prefab

fireplace. Several of the homeowners stated that the improvement of

prefab fireplaces in recent years have made them appealing. Only 20

percent of the apartment dwellers indicated a preference for a prefab

fireplace.

Builders surveyed indicated that recent trends in preferences for

fireplaces have increased [National Consumer Survey on Housing, 1978].

In the price range of the study ($38,500), the prefab fireplace is the

only logical choice because of cost feasibility.


Response 21f. Exterior option aluminum soffits.


0 1 TOTAL

A 21 5 26

H 13 16 29

TOTAL 34 21 55

Chi-square = 7.503 D.F. = 1

Prob. = 0.0062









Chi-square tests for differences in preferences for six exterior

options yielded only one significant difference. Homeowners, possibly

because of their experience with exterior maintenance, chose to include

aluminum soffits in their preferences. This option is becoming very

popular with buyers interested in reducing maintenance [Consumer Prefer-

ence Survey, 1977].


Optional Feature Tradeoffs


In the personal interviews, the respondents were aware of the

price of each choice they indicated; however, they were permitted to

choose any of the optional features preferred in a single-family home.

The maximum price of $38,500 made it necessary for many of the

respondents to reduce their choices of some of the optional features

they had chosen. Although few individual optional features were

consistently deleted to reduce the total price of the home, the study

did identify types of optional features that the respondents would

give up to reduce the price.


Landscaping

The respondents were given three choices of landscaping packages

ranging from basic to complete. Many respondents at first indicated

a preference for landscaping that would require little or no additional

work by them upon moving into the new home; however, to reduce the

price of the home, almost 50 percent were willing to reduce landscaping

expenditures by $500 to $850. The respondents indicated that additional

landscaping could be installed by themselves after occupying the new

home.









Insulation and Energy Options

Respondents indicated that their preferences for additional

insulation and energy options were the result of media exposure and

product advertising. When confronted with a maximum price constraint,

the respondents were willing to give up several of the energy options

chosen first. They rationalized this tradeoff by stating that they did

not anticipate living in a home long enough to receive energy savings

sufficient to cover the original cost of the energy options.

Additional insulation was not as susceptible to deletion, however.

Respondents seemed to be aware of the potential savings of electricity

and value at time of resale. Fewer than 10 percent of the respondents

deleted this optional feature to reduce the price of the home.


Exterior Options

Three exterior options regarding outdoor patios were offered.

The respondents were typical of the young "do-it-yourselfers." Many

of the subjects indicated that for the price of these options, they

could later install a patio if desired.

Aluminum soffits were another exterior option chosen. Over 80

percent of the respondents who preferred this option did not delete it

to reduce the cost of the home. The cited reduced maintenance as the

reason for choosing this option.


Space Tradeoff

Respondents indicating the need for a walk-in closet were asked

to give up comparable area in other parts of the home. Subjects were

willing to give up space in the master and secondary bedrooms, but not

in the kitchen/dining or living area.









Lot Size

More than 85 percent of the respondents first indicated the two

larger lots as their preferences; however, fewer than 20 percent

elected to reduce lot size to lower the overall cost of the home.


Heating and Cooling Options

More than 95 percent of the subjects surveyed indicated a prefer-

ence for central air and heat; yet none of the respondents was willing

to delete this optional feature to reduce housing costs. The second

most popular heating and cooling option (prefab fireplace) was

deleted by almost 30 percent of the respondents when it was necessary

to reduce the price of the home.


Demographic Characteristics


The subjects surveyed in this study were from the Gainesville,

Florida, metropolitan area. Because of the relatively high percentage

of students in the sample, a test was completed for differences in

preferences between students and non-students. No significant dif-

ferences in preferences were found when tested at the 0.05 level.

The primary objective of the study was to test for differences

in preferences between the two groups surveyed. A comparison of the

demographic characteristics of the two groups indicated that no

significant differences existed. It is concluded that no response bias

was caused by demographic characteristics of the respondents.

Previous studies nave attempted to segment the market using

various demographic characteristics (see Chapter II). These studies









identified age, income, number of children living at home, and previous

home ownership as the most important explanatory variables.


Age

As previously stated, the sample was drawn from families living in

homes costing approximately $38,500 and families in apartments. Due to

the age composition of the population sampled, 93 percent of the

respondents in the sample were under 35 years of age. This skewed

distribution of ages should not create any response bias related to

the testing of preference differences; however, response bias would

appear if predictive behavior of potential consumers was attempted.


Income

Income was not considered to be an important demographic character-

istic in this study. It was assumed that the population sampled was

homogeneous and that the variable of income would not create response

bias. In other studies the researchers attempted to use responses from

subjects to predict consumer behavior in housing purchases. In these

studies, the question of affordability of the optional features chosen

was not tied to income. Since the objective of this study was to test

for preference differences between consumer groups using alternative

methodologies, income, which was relatively equal for the respondents

(except students), was not considered to be a factor that would affect

the results of the study.


Number of Children Living at Home

The respondents in the study had from zero to five children living

at home. Since the average number of children living at home was two






79


for both samples of the two groups tested, this demographic characteris-

tic did not appear to be the source of any response bias.


Previous Home Ownership

A test of preference differences between persons who had previously

owned homes and those persons who had not previously owned homes yielded

significant differences for several optional features. The differences

are summarized in the previous section entitled Test of Differences

Between Homeowners' and Apartment Dwellers' Preferences. The findings

of the study are consistent with those of other studies.

This element of the study has some predictive potential for

researchers who may be attempting to segment markets by first-time

buyers. In the price range of the study, optional feature preferences

were ascertained for these two groups.














CHAPTER V

CONCLUSIONS AND RECOMMENDATIONS


Conclusions

Research in the housing industry has sought to identify consumer

preferences related to optional features. The literature review

revealed that efforts to do so have been incomplete. The purchasing

power of the consumer has not been considered an important factor in

determining the validity of the preferences indicated. Due to the

absence of price of optional features in the preference studies,

researchers may have obtained misleading consumer preferences.

Feature preference studies conducted by building trade journals,

building associations, and real estate development firms have utilized

two data collection techniques. Building trade journals have used a

mail-out questionnaire to a large sample representative of each major

region of the country. These families were surveyed to ascertain

preferences regarding house design, optional features, neighborhood

quality, and location. However, inconsistencies were found because

consumers were not aware of the costs of the optional features chosen;

additionally, the features, size, and price of the home chosen were

not within the affordable price range of the respondents, as determined

by their total family income.

Building associations and real estate development firms have

surveyed potential consumers at several subdivisions. The responses









obtained were to be used by builders to respond to consumers' wants and

desires. The researchers admitted that an association bias may result

when potential buyers are surveyed upon leaving a subdivision; they

are influenced by what they have seen in that development.

Research of these types (mail-out questionnaires and point of sale

surveys) often give subjects an unlimited list of optional features to

choose from. Some constraints must be used if the preferences indicated

are to be useful to designers and builders.

This study sought to test for differences in preferences of

optional features of homes between one group who had knowledge of the

price of each choice and the other group who had no knowledge of the

price of each choice. Significant differences in individual prefer-

ences were found using Chi-square tests. Many of the differences of

individual preferences resulted in a higher total cost of the home.

Significant differences were found in several areas of optional

features. The significant difference in preferences between the two

groups resulted in a higher price for housing. The group of respondents

unaware of the price of each option had significantly different prefer-

ences of optional features that resulted in a price of almost $4,000

higher than the average price of the home chosen by the group with

knowledge of the price of each option.

Several preferences for optional features that accounted for the

overall price differences include the following:

(1) Masonry fireplace preferred over the prefab fireplace

resulted in $1,000 added cost.

(2) Higher quality shingles preferred over standard shingles

added approximately $500 to cost.








(3) Insulated windows chosen over standard windows added $500 to

total cost.

(4) Solar water heater added to the home would cost $1,500.

(5) Choosing largest lot would add $1,000 to the price of the

home.

These major differences in preferences resulted in an observable

difference in the mean price of the homes of the two groups.

Using the gross total price of the homes of each group, the means

were calculated and tested for significant differences. The hypothesis

that the means of two samples were not significantly different was

rejected at the 0.05 level with a Z-value of 9.755. This indicated

that significant differences in individual preferences resulted in a

different overall price of homes.

Based on these two tests, the hypothesis that there is no dif-

ference in preferences of consumers with knowledge of the price of

each optional feature and consumers without knowledge of price can be

rejected. Implied with the rejection of the hypothesis is that con-

sumers with price data indicate different preferences because of

better information.

Researchers are often disenchanted with market research in

housing because consumers, based on surveys, indicate preferences for

a type of home and then purchase something different. As explained

in the literature (Chapter II), this may occur because consumers want

or desire one type of house with certain features, but are unable to

purchase such a home within their budgetary constraints. These con-

sumers reformulate their preferences subject to their income constraints.









The results of this study indicate that consumers, constrained by

a maximum price, can maximize the housing space and optional features

available. Consumers' preferences can be obtained yielding more useful

information, if the optional features chosen are within the budgetary

constraint of the respondents.

The primary objective of the study was to determine whether dif-

ferences of preferences occurred between the two groups surveyed. The

study accomplished this by providing evidence that there is a differ-

ence in the preferences of the two groups (group with knowledge of

price and without knowledge of price).

Other secondary conclusions can be drawn from the study. The

preferences of consumers who have previously owned a home were compared

with those of consumers who have not previously owned a home. Signifi-

cant differences were shown to exist.

First-time buyers represent a large number of potential consumers

in the housing market. A technique, such as the one employed in this

study, would allow researchers to determine the areas of preference

differences for optional features in housing.

The use of optional feature tradeoffs was attempted in this study.

Constrained by a maximum affordable price, consumers must be selective

in their choices of optional features in a home. When consumers are

required to reduce the expenditure for a home, builders need to know

the relative importance placed upon various optional features. By

knowing this, builders may delete these items in reverse order of

importance to reduce the total cost of the home.








Recommendations


This study was designed as a pilot test of the hypothesis that there

is no difference between preferences of consumers with price knowledge

and those without price knowledge. The optional features included in

the survey were those generally available in homes of the price range

of the study ($38,500). For this study to be replicated in other price

ranges, optional features that are cost feasible in the price range

must be identified and the costs specified.

To insure that respondents are aware of all optional features, a

list should be made available at the beginning of each interview.

This modification in the graphic display kit became evident because

many respondents were startled to find near the end of the interview

that the optional features chosen did not include central heating and

air conditioning. By including a listing by major categories of the

optional features, the respondent would be aware of the options

available.

Another area that caused minor problems was the unit pricing of

floor covering. Some floor covering costs were stated using the price

per square foot while others are stated using the price per square

yard. To remove potential confusion for respondents, all unit costs

should be stated in the same unit of measure.

This study was only a pilot study; therefore, the process of using

the two survey methods should be replicated using a larger sample.

The transferability of the survey instrument to other areas would

require the researcher to change the optional features available to

the respondents to reflect local availability and desires.









To use this graphic display kit and accompanying survey form, one

would need to determine the maximum price of the home and then deter-

mine the size of the typical homes in the price range. Using the

maximum price of the home as a guide, the researcher would identify

the optional features that would be cost feasible. Using the price of

each option, consumer preferences based on responses could be obtained.

In the event the maximum price is exceeded, the respondents would be

asked to delete optional features until the desired price is obtained.

The final recommendation is for use of the graphic display kit by

builders. Frequently, builders sell homes to families and then begin

construction. At the beginning of the sale, the builder using the

graphic display kit, could show potential home buyers the various floor

plans available, room sizes, and optional features available. At the

conclusion of the process, not only would the builder know which

optional features the consumer preferred, but the home buyer would be

aware of the price of each item and could more knowledgeably choose the

features desired.





























APPENDIX A

LETTERS TO HOMEOWNERS AND APARTMENT DWELLERS












COLLEGE OF BUSINESS ADMINISTRATION UNIVERSITY OF FLORIDA GAINESVILLE 32611
REAL ESTATE RESEARCH CENTER 904 392-0157
March 29, 1978










The Real Estate Research Center at the University of Florida is conducting
a study to learn how recent home buyers believe their homes could be
constructed better or improved. We would like to schedule a brief inter-
view with you in your home to learn of your opinions.

The results of this survey will be used by builders and developers of the
area as they try to respond to home buyers' preferences. Your experience
as a home buyer and homeowner is very valuable to us in this research
project.

The actual survey is relatively brief and can be completed in the home.
However, it is important that if you are married that both spouses be
present when the survey is completed. This assures us that the responses
are a "team effort" and that the preferences reported reflect a combina-
tion of both spouses' desires and wishes regarding housing preferences.

We will telephone you during this next week to arrange a mutually conve-
nient interview time. The interview should last approximately 20 minutes.
In the meantime, if you have any questions, please call me at 392-0301.

Sincerely,


L. M. Dyson, Jr.
Research Associate











COLLEGE OF BUSINESS ADMINISTRATION *
REAL ESTATE RESEARCH CENTER
March 29, 1978


88





UNIVERSITY OF FLORIDA GAINESVILLE 32611
904 392-0157


The Real Estate Research Center at the University of Florida is conducting
a study to learn how people currently living in rental housing believe
single family homes could be constructed better or improved. We would like
to schedule a brief interview with you in your apartment to learn of your
opinions.

The results of this survey will be used by builders and developers of the
area as they try to respond to first-time home buyers' preferences. At
some point in the future you may be one of these home buyers.

The actual survey is relatively brief and can be completed in your
apartment. However, it is important that if you are married that both
spouses be present when a survey is completed. This assures us that the
responses indicated are a "team effort" and the preferences reflect a
combination of both spouses' desires and wishes regarding housing prefer-
ences.

We will telephone you during the next week to arrange a mutually conve-
nient interview time. The interview should last approximately 20 minutes.
In the meantime, if you have any questions, please call me at 392-0301.

Sincerely,


L. M. Dyson, Jr.
Research Associate





























APPENDIX B

SURVEY FORM TO ACCOMPANY COST DATA
AND GRAPHIC DISPLAY














0 C
0 0

-o
)C 4-'
S- 0
0~-- 2


no 0 4-
0c0 0 0

- 4- L)


LaD4--'
o O +. C
a)1 a 0 10
E 0
4-4..


2 4-0








401
CL 0
> 0.- 0 r









S- 3 C
2 S...0






()j (u


2 2 4-' 4
S..0 2








2r / 4-'










-0 0
.200















CL 04-
r- n u o,
















LU~








M._
EE~
U 4 0 M


. r0 .
)- 0
















3 -- a)Ua
mr 4- -
















EC < 0) .
0'0
S 4-
0~ 2




























0-
*-0 0











4-mu
4 C > E

S- *0- i-
"21 0.
>i(Ur i-Q
ro ro -






0.-'- 5
4 M0 m
Q-2 0


CL0 0 2
'r2.00








-c 2 0 0

5.- 02i/
a1 >12 2O<
LU 0
'a 2e -
04-.~ 2 2







03 *- 0>


2 24- 0f 0O

c 2c5. >,
o e










04 0. .r0< (
4-0 50
E~










230- 0C

04- 0 -^ -



0f0 2
D*- 2 2,r 0-
0 4- 2.0 2
u o o c
















2 ..2- 0*
( 0 U (J
0J~ 1=E* r




r r- f) .0 (

~V) S- U 4

c CLJ3 0 l-


S-
2
-r-

(UI
0

CL
V)

0
0
S..

0
=3
0.












.0
4-
2
2
















co
0



r
3






"5
,.0















E
2
0
0















= 4
I-







2
.0

.0


4- *-
nO0 0 C o (' cn

LU >12<

0 20tl 0. 0

LU 0.0C
QT = 4-'


0

0

0



-o


Sc


='4.0 (o < ) co
LoB


0 E

LU 0
a: n=


4-'
22

((0

0
4-4 0
S- -


aj
23 m4-



mm
r-~
41
.0 2









C)T
*.- SQ

24-'
0 0n











ucue
.0 0 f0









2 C

0
a0r
0 4u-



f0 / cJ
5- o-r



4-'

4-'- -0
L 0 m



V)> a,
0 0.
0l+J 0 .0
t 4l Q- (U
0UC S
2^ 25 t~(








91











Li


4-




0
'a
0

0



E U- LLU U
0 (r) S VS V
0
-o
S. 0- L.) l) L.)

.0
S- C) C 0 O Ci



5--
a, CM CM '- -O
aj


Sx x x x


C'U
ES 0 i- CMj CM)
*i"a, r
U N

CL m 4
0
V a
Lii .C0
ce: 5


5-
.C
4- M
0
4-' 0


Em

,?
4 (

E-=
nci



.0

VU 41
*4-' (1)




ci 4-'O V
a0 -c






o o
00
E











00 0 c ci .
.03 000:



0
cau



(I4- G
I n t U_ LL L- L-








S0 0 0) 0i CM
(0 o- o i- CM m ta

E0E
0 0 0 -- C\ CN C\





0. 0 r r
S U 0}


O^ -o cc CQ u



Li'- 0~
rI 1 4- (U
Ct: u- /


LI) EnI
0






a)
tU LL- LU Li Li
C V ( V 0) C S
0 U) tO In c


0
-0
US1 C.) a,-


m Q < C



8 E
CL
uj, o ~ 0 L
occ i








92







TS
C)

LU











-j
(J>





0
C-)
LU

:3.
F-


C-)


41

4
40 0
-C

S-'
(0 (0-*-




4.'.( w 0
(U >i 0 0 L



>' -- 4-
0
Ln
0I 0 D 'O N
*r- (E A C: C'
04 0( CO ) *r (



(U:3 :3 c
0--
"CU
S (U






C:C) S S L 1L
Q 41
(0 (V _
*r 13 OJS OJ '0





= ja > ^ -^ ^
,u,
0r .' ( / i- C. (0
i 5<
=. LU U o


0- -- .- -
5.(0 "S C')
C. C
LO UJ0~ 0



EJC-C (U03
CO >- C: >


LD LO nO
LO 0? LO C? 0
C'J t- CJ r- C
S-n co CM 0t n
3: U 0 C -r U W.
j aj S- u S-
.S- CL: c L U- U. U.
a) '0 (n O (A ( V

U, *i
SL.4--C:in 0 0 0 CO CO


4 0 X
(0N>,L C CO CO 0 0


4 C -U 00 co CM Cn
c 4-'
-^+- Q. co co cr C
4-'V( U 0
0l~~ vl rt fO <
i =m m a (
0O 3 0
41 *i- 0 0 0




C -0 w
,( ( 4 (

04.'(n E 2 2 2 o

41 C C r X X -
i/l-0 ~ :3 :33 i 0 i "



0 s- U w
f-' u -o i- o -: o 5
'D IraC <-S C



V (DJ5 (0 m0 '
S* CLU

CLTC- ) 1 1
-i-.c u u t/ E E E E

4i- > +.vl 1 C: *4*- *" -- _
(A -4 ( (00(0 0.S 2
*r- 0















































































o 0 0
o 0 CCMS LO
C(J LO to-





to
to 0
o c
0 a, a
to 0

0 to
a) M


0,



tco to
(U (0 t0 o

V) (U > (0
or f0 (U "
0 C t/ 2

eT CQ 0


EE
t- 0 0t
to toO L
S- 0 SCS 0




t,-
(v s-r
+4- r )





umua
'ow
m0V CD CD
to to 0..-





OS.- 0
U Q)
a) 0
4-' to to





004-
;C S.- 41
r CL40






*iL ( 3 Cv. (U
S. o anc'
Co 5- 4- 4--' 1-
C) ( 4- o-m o0
Oo Ui

0 (D40

0 0.4-' 0 to
reo +4.0-c
0.0 to 04-'






Q0 0o 01' 5-+

0 -)4-' t- /-*
5/ 4' 0J to to

co to> '- .0ro 0.
oto Otou


Q-tl U, arr






VS to to 4-' 4-' to
U.S J= 0.0
Q; m --i i -


4-
U)
0
S.)

0'

5.-



0
(-S


*a
to

--t

0-





o.'-1
r-











>.4-'
*to 0.
0
(0 r


00





0 4-'
to

0 O
(~0
S.--




>to
tow
tou
aoo
ce '




o~o
(T U




0. to
to.- *o
US .0.
t/ *'- 4-'
LU^:^
Q;[ *-


LL- L
C) COi









X 0
CO C
to ro

0 0


0 0

0 0


LL. U..
L L
L/) V)







50 Wo
4D CL


0

0 to
0 0
to
4-' 0.

*- c
-S's
m
to 4-'
5- to
Co U)
Qu LU








94




5


LU










V)
w

~-J


4-




F-

















_0f




















0
4-)
LU

F-t

-J


























w w)
ow


C> w 4- M

0LU 4 -
4"







C : c
4r tfUCL
C) 4/) 4/) (/0 4/) 4/) 4/) 4/) 4/") 4/) 0 > (0
-e 0 *r U(U
0 oo Csi 0 4/ 4/ 4/) 4) C 4.4-' 4-44..
^ CM 4DO 0US i -)~ C-)-C)t O t
o r- r- C) C








LO S-Y
41 w C) c a) c 42
m) C m 4- CU CJ c



=) CD w ) wa
C) .-j C)0C
04" L' L Y~ e


4/) w 00
(0 -- Q0jC) 44-C )
-J *C4-4 (U i)




0 *r1 CD 4. r4- cc
CC CD C) w- ')0C -


c o w 0 w





V)L- 41 V)4 u4
0 C34-o CD 0
4-'- ^ C- fO : ')C )> C) -- 4


CL ^ 40 4-) 0 w=

V) -0 vi i S +-


ui -C U a-
4-' w Ln W= i=- 40- 4-' 4
C) )< )< 4( 4- 01 0: 4Q" tC)-0 ( 0 0 C) C



(U p_ o- ^ O c cr cuca D- 4
u F F- O LU C) 0.0j3 C) 0 c0 C
i-*- O CM CMJ ^ LU LU C4: 0) XC-.- CM 4-'0 0 0 0 4-
^-.0 ^- - C LO/) 4/ 0^ 4-' &-T->4 o 4i NM 0
S 0 0 0 LU >- '^-
LU.i J1 .J .J 0) 00E:C LU O C)
tC) ^. 40 4 0J 0_ C)j CC .0C)> C: C))- 4 0
04)4 l) 3^ LU Z -0.OC) 0C)4-'
0.C f Ot a LU 4-) F 0 C) 4-' .4 ) r
4/)C) .. F- >^ C) 4/) 4-)J 4)4.0 V) C)4-.
LU,- 4-Ll -. >[. S: LU c)0 L3 U000i-4-