|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
STRENGTHENING THE VOICE AND FARM BUREAU FOUNDATIONS: EXPLAINING
LOCAL BOARD PERFORMANCE AS RELATED TO A PROFESSIONAL DEVELOPMENT
PROGRAM FOR A NONPROFIT MEMBERSHIP ORGANIZATION
ERIC K. KAUFMAN
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
O 2007 Eric K. Kaufman
To my wife, Shevon;
and to a bright future for my son, Ethan.
I would like to thank my supervisory committee: Dr. Glenn Israel, Dr. Rick Rudd, Dr.
Nick Place, Dr. Ed Osborne, and Dr. Elizabeth Bolton. I appreciate all of your insights and your
contribution to the quality of my educational experience at the University of Florida. In addition,
I thank Dr. Hannah Carter for providing some of the early inspiration for the research and
leadership development programming with Florida Farm Bureau. You have been a valuable
mentor throughout my graduate educational experience. I thank the Florida Farm Bureau
Federation for their sponsorship and support of my research efforts. I particularly appreciate the
leadership and guidance of Pat Cockrell and Ray Crawford, but I am also indebted to many of
the staff members and volunteer leaders for their assistance in completing a wide variety of tasks
associated with my research. Finally, I would like to thank my fellow graduate students (past
and present) in the Department of Agricultural Education and Communication at the University
of Florida. You have made the experience bearable, and often times even enjoyable. Thank
TABLE OF CONTENTS
ACKNOWLEDGMENTS .............. ...............4.....
LIST OF TABLES .........._.... ...............9.._.._ ......
LIST OF FIGURES .........._.... ...............11._.._. ......
AB S TRAC T ............._. .......... ..............._ 12...
1 INTRODUCTION TO THE STUDY............... ...............14.
Background of the Study .............. ...............14....
Goal Setting Theory ............... .. ...............15...
Nonprofit Organizational Effectiveness ................. ...............16........... ....
Farm Bureau's Leadership Needs .............. ...............17....
"Strengthening the Voice" Program ............ ......_ ......._ ...........1
Overview ..........._.._. ........ __ ... ...............19..
"Farm Bureau Foundations" program component .............. ....................2
Problem Statement ..........._.._. ........ ...............24.....
Purpose Statement .............. ...............25....
Research Objectives............... ...............2
Significance of the Study ..........._.._. ........ ...............25....
Lim stations ..........._.._. ........ ...............26.....
D efinitions .............. ...............27....
Summary .........._...._ ......_. ...............28.....
2 REVIEW OF LITERATURE ..........._.._. ........ ...............34....
Theoretical Frame .........._..__ .. .. ...._.. ...............34.....
Goal Setting Theory Defined............... ...............34
Goal setting and planned behavior .........._...._ ......_. ........_...........3
M echani sm s ................ ...............36........... ....
Content .............. ...............36....
Intensity ................. ...............37.................
Other moderators ................. ...............38.................
Demographic variables............... ...............3
Adult Education ................. ........... ...............41.......
Andragogy and Adult Learning ................. ...............41................
M otivation to Learn ................. ...............42........... ....
Program Planning and Evaluation .............. ...............44....
Andragogy in Practice Model ................. ...............44................
Conceptual Programming Model .............. ...............46....
Program Evaluation ................. ...............47.................
Overview .............. .... ...............47.......... ......
Return on investment .............. ...............49....
Leadership Development in Context .............. ...............50....
Rural Leadership Development ................. ...............50.____ ......
Agricultural Leadership Development .............. ...............52....
Grassroots Leadership Development............... ..............5
Nonprofit Organizations ............_. ...._... ...............56....
Overview .............. ...............56....
Leadership ............... ..... ...............58.
Florida Farm Bureau Federation............... ...............6
History .............. .. ..... ...............61
Mission, Vision, and Goals .............. ...............62....
Local Farm Bureau Units .............. ...............63....
Local Boards of Directors .............. ...............63....
Field Staff ............... ...............64.
Sum m ary ................. ...............65.......... ......
3 METHODS .............. ...............69....
Research Design .............. ...............69....
Populations ................ ............ ... ... .. .. ............7
Participant Selection: FFBF Local Board Members .............. ...............71....
Participant Selection: FFBF Field Staff ................. ......... ......... ...........7
Instrumentation ............... ....._ ...............74....
Local Board Practices Survey .............. ...............76....
Best Practices Goal Survey .............. ...............76....
Field Staff Perceptions Survey .............. ...............77....
Data Collection Procedures .............. ... ...............78..
Institutional Review Board Approval ................. ............ .... ............... 78....
Collection Procedures with FFBF Local Board Members .............. .....................7
Re-group and additional mailing ................. ...............82................
Unit response rate ................. ...............83...............
Additional data from FFBF .............. ...............84....
FFBF Field Staff. ...._.._................. ........_.._.........8
Data Analysis Procedures ...._.._ ................ ........_.._.........8
Unit Non-Response .............. ...............86....
Item Non-Response .............. ...............89....
Dependent Variable Indexes................. ...............9
Individual-level dependent variable ...._.._ ................ ........_.._ .........9
Board-level dependent variable ....__. ................. ...............91......
Independent Variable Indexes .............. ...............94....
Goal attributes .............. ... .... ...............9
FarmFFFFF~~~~~~~FFFFFF Bureau Foundations program participation .............. ....................9
Weighted goal theory constructs ........_........_. ..._........... ............9
Board demographics................ .................9
Analysi s for Obj ectives One Through Three ....._.__._ ........_. ....._._. ........9
Analysis for Obj ectives Four and Five ..........._...__.......... ....._. ..........9
Goal -related hypotheses .............. ...............98....
Correlational analy sis............... ...............98
Regression analysis .............. ...............99....
Path analysis............... ...............10
Summary ............ ..... ._ ...............103...
4 RE SULT S ............ ..... ._ ...............109...
Obj ective 1: Describe Demographics ............ .....___ ...............109
Obj ective 2: Measure Engagement Level ....__ ......_____ .......___ ............1
Obj ective 3: Measure Goal Attributes .............. ...........__ .. ...............113
Obj ective 4: Explain Individual Performance on Best Practices............. ..__.........__ ....1 14
Obj ective 5: Explain Local Board Performance on Best Practices ............_.. ..............1 18
Sum m ary ................. ...............123..............
5 SUMMARY AND DISCUSSION .............. ...............140....
Summary of the Study .............. ...............140....
Statement of the Problem ............ ........... ...............140.....
Purpose and Obj ectives ............ _...... ._ ...............141..
M ethodology ............ ..... .._ ...............141...
Findings ........._ ........ __ .. ......._ ............14
Objective 1: Describe demographics............... .............14
Objective 2: Measure engagement level .............. ...............143....
Objective 3: Measure goal attributes.................. ................. 143
Obj ective 4: Explain individual performance on best practices ...........................144
Obj ective 5: Explain local board performance on best practices.............._._.........144
Conclusions and Discussion ................._ ...............145........ .....
Obj ective 1: Describe Demographics ........._.._.. .......... ...._.. ..........14
Obj ective 2: Measure Engagement Level ....__. ................. .. .....__ ..........14
Obj ective 3: Measure Goal Attributes ............... .. ........_.._ .......... ..........14
Obj ective 4: Explain Individual Performance on Best Practices. ........._.._... ........._.....151
Obj ective 5: Explain Local Board Performance on Best Practices .............. .... ...........154
Im plications .............. ...............158....
Goal Setting Theory .............. ....... ..............15
"Strengthening the Voice" Program Theory .............. ...............158....
Recommendations................. ... .. ........5
Suggestions for Additional Research............... ...............16
A BEST PRACTICES GOAL SHEET .............. ...............163....
B WORKSHOP FOLLOW-UP LETTER ................. ...............164....._.._....
C WORKSHOP FOLLOW-UP NEWSLETTERS .............. ...............165....
D LOCAL BOARD PRACTICES SURVEY .............. ...............170....
E BEST PRACTICES GOAL SURVEY ............ ......__ ...............178.
F FIELD STAFF PERCEPTIONS SURVEY. .....__.....___ ..........__ ............8
G COGNITIVE INTERVIEW PROTOCOL .............. ...............191....
H PRE-NOTICE LETTER TO STUDY PARTICIPANTS .............. ..... ............... 19
I INITIAL QUESTIONNAIRE MAIL-OUT LETTER ................. .............................193
First Questionnaire Mail-out Letter for Program Participants ................. ......................194
First Questionnaire Mail-out Letter for Program Non-Participants .................. ...............195
J THANK-YOU/REMINDER POSTCARD .............. .....................196
K REPLACEMENT QUESTIONNAIRE MAIL-OUT LETTER .......__ ......... .......... .....197
Replacement Questionnaire Mail-out Letter for Program Participants .........._... ..............198
Replacement Questionnaire Mail-out Letter for Program Non-Participants ................... .....199
L TELEPHONE INTERVIEW PROTOCOL FOR FOLLOW-UP WITH NON-
RESPONDENT S .............. ...............200....
M E-MAIL TEMPLATE FOR PRE-NOTICE TO FIELD STAFF............_._. ........._._.....203
N INFORMED CONSENT LETTER FOR FIELD STAFF .............. ..... ............... 20
LIST OF REFERENCES ................. ...............206................
BIOGRAPHICAL SKETCH .............. ...............218....
LIST OF TABLES
1-1 Farm Bureau Found'ations workshop best practices by obj ective ........._._. ..........._......33
1-2 FFBF Farm Bureau Found'ations workshop dates and participant numbers by district....33
3-1 FFBF Local Board Member Research Study Sub-Groups .............. .....................0
3-2 Item non-response by population group .............. ...............107....
3-3 Usable survey response by population group .............. ...............107....
3-4 Goal specificity construct: Factor loadings from the "Best Practices Goal Survey".......107
3-5 Goal difficulty construct: Factor loadings from the "Best Practices Goal Survey" ........107
3-6 Goal commitment construct: Factor loadings from the "Best Practices Goal Survey" ...108
3-7 Goal self-efficacy construct: Factor loadings from the "Best Practices Goal Survey" ...108
3-8 Goal performance construct: Factor loadings from the "Best Practices Goal Survey" ...108
4-1 Demographics of FFBF Local Board Members and population sub-groups ................... 125
4-2 Ratings of individual and board practices among FFBF Local Board Members and
population sub-groups............... ..............12
4-3 Perception of engagement level (board and individual) in Farm Bureau related
practices in comparison with other six month periods............... ...............128
4-4 Farm Bureau Found'ations' participants' ratings on questions related to the goal
setting process ........._.. ..... ._ __ ...............129....
4-5 Farm Bureau Found'ations' participants' ratings on constructs associated with goal
setting ................ ...............130....... ......
4-6 Board-level variables' correlations with ratings of personal performance on local
Farm Bureau practices .............. ...............13 1...
4-7 Individual-level variables' correlations with ratings of personal performance on local
Farm Bureau practices .............. ...............13 1...
4-8 Goal-theory constructs' correlations with ratings of personal performance on local
Farm Bureau practices .............. ...............13 1...
4-9 Inter-correlations between selected independent variables at the individual level of
4-10 Regression models for individual performance: Explaining average rating for
personal performance among best practices using variables associated with goal core,
moderators, and mechanisms ................. ...............133._._.. .....
4-11 Standardized regression models for individual performance: Explaining average
rating for personal performance among best practices using variables associated with
goal core, moderators, and mechanisms .............. ...............134....
4-12 Correlation between ratings of board performance on local Farm Bureau practices
and other ratings of performance ........._._.._ ...._... ...............135...
4-13 Correlation between average board member rating of board performance on local
Farm Bureau practices and board demographics ........._.._.. ...._.. .........._.......135
4-14 Correlation between average board member rating of board performance on local
Farm Bureau practices and FarmFFFFF~~~~~~~FFFFFF Bureau Foundations workshop attendance ................136
4-15 Correlation between goal-theory constructs and ratings of board performance on
local Farm Bureau practices............... ...............13
4-16 Regression models for board performance: Explaining aggregate board member
rating for board performance among best practices using weighted goal theory
constructs, program participation, and board demographics ................ ........_.._.. .....138
4-17 Standardized regression models for board performance: Explaining aggregate board
member rating for board performance among best practices using weighted goal
theory constructs, program participation, and board demographics ............... .... ........._..139
LIST OF FIGURES
1-1 Essential elements of goal-setting theory and the high-performance cycle. ....................30
1-2 Florida Farm Bureau Federation policy development and implementation process. ........30
1-3 Process model for the Florida Farm Bureau Federation "Strengthening the Voice"
program ................. ...............3_ 1....._.__....
1-4 Florida Farm Bureau (FFB) "Strengthening the Voice" program logic model. ...............32
2-1 Theory of planned behavior. ...._.. ....... ....._._._. ......._.._ ........_........67
2-2 Andragogy in Practice Model. ............. ...............68.....
3-1 Research procedures timeline. ............. ...............105....
3-2 Number of individual respondents from each local board ................. ......................106
3-3 Hypothesized relationships between goal constructs and performance...........................106
4-1 Correlations between goal constructs and personal performance graphically
displayed in the hypothesized model ................ ...............124........... ...
4-2 Correlations between goal constructs and local board performance graphically
displayed in the hypothesized model ................ ...............124........... ...
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
STRENGTHENING THE VOICE AND FARM BUREAU FOUNDATIONS: EXPLAINING
LOCAL BOARD PERFORMANCE AS RELATED TO A PROFESSIONAL DEVELOPMENT
PROGRAM FOR A NONPROFIT MEMBERSHIP ORGANIZATION
Eric K. Kaufman
Chair: Glenn D. Israel
Major: Agricultural Education and Communication
Agricultural leadership development can serve both the agricultural industry and the
communities where agriculture is found. Goal setting theory, which is based on the premise that
conscious goals affect action, is an underused framework for improving agricultural leadership.
Aspects of goal setting theory were applied in a leadership program conducted by the Florida
Farm Bureau Federation for its local board members. That program provided a context for this
study, and the program's target audience was the study population. The study's purpose was to
explain performance, as related to professional development for local boards of a nonprofit
organization. The research was intended to identify factors that explain the engagement level of
local board members in a multi-level, membership organization.
Study participants tended to report that they were "mostly" engaged in the local Farm
Bureau practices promoted in the program. Although program participants' overall use of best
practices was not significantly different from non-participants, participants reported significantly
higher engagement levels for specific practices: Strengthen the 'Voice of Agriculture' by
getting involved in statewide committees"; "Plan for policy implementation challenges before
they arise"; and "Consult others, including adversaries, for input and feedback on policy ideas."
Overall, program participants believed the program helped them to become more effective
Farm Bureau members and leaders. However, program participation alone did not explain
differences in individual and board performance. Instead, gender, education level, meeting
attendance and perceived differences in the goal setting process explained differences in
performance. Regression models including these variables explain up to 21% of the variance in
individual performance and 24% of the variance in board performance.
This study has implications for professional development programming and application of
goal setting theory. The researcher recommends that the program continue, provided that
implementation is standardized and emphasis on goal setting is increased. Additional research is
needed to further explore the application of goal setting theory in programming for volunteer
leaders. An important research focus is improved measures of goal theory constructs and
volunteer board performance. Research also is needed to better understand the differential
effects that gender, education level, and meeting attendance have on board performance.
INTRODUCTION TO THE STUDY
Agricultural leadership is needed today more than ever before. With the agricultural field
becoming more specialized and increasingly challenged, the future success of the industry is
dependent upon local leaders to guide efforts for advocacy and change (Diem & Nikola, 2005;
Horner, 1984; Howell, Weir, & Cook, 1982; W.K. Kellogg Foundation, 2001). Even beyond the
industry, though, the fate of many agricultural communities is also at stake. Agriculture has
traditionally played an important role in most rural communities, and many of these communities
are falling behind their urban counterparts in many areas of quality of life (W.K. Kellogg
Foundation, 2004). Without capable leaders, these communities are "prone to inertia, decay, and
manipulation" (Hustedde & Woodward, 1996, p. 1). Leadership, by definition, inspires vision
and hope for the future (DeRuyver, 2001; Kouzes & Posner, 2002; Moss & Liang, 1990).
Agricultural leadership development holds the opportunity to serve both the agricultural industry
and the communities where agriculture is found.
The phrase "agricultural leadership" lends itself to a broad range of ideas, topics, and
programs. Research on the topic reveals that related subj ect areas include leadership programs
for rural youth, leadership education programs in colleges of agriculture, adult agricultural
leadership programs, governmental agricultural agency positions, and general rural community
development efforts. Among these topics, all require additional research to further the field and
scholarship of agricultural leadership (E. Osborne, 2007; Rudd, Stedman, & Kaufman, 2004).
Background of the Study
This study investigated the usefulness of goal setting as applied in the context of a specific
leadership development program conducted by the Florida Farm Bureau Federation.
Goal Setting Theory
Among the many definitions of leadership, Locke (2003) defines leadership as "the process
of inducing others to pursue a common goal" (p. 29). This definition has emerged from decades
of research by Locke and others on goal setting theory. Locke suggests that the value of goal
setting extends to a wide variety of applications. However, past research on goal setting theory
has focused primarily on motivation in work settings (Locke & Latham, 2002, p. 714). Although
some agricultural leadership takes place in work settings, many of the contexts for agricultural
leadership are in community and volunteer settings. Research is needed to determine the
appropriateness of goal setting in these contexts for agricultural leadership. The use of goal
setting in these contexts is not entirely novel, but the formal application of goal setting theory is.
Goal setting theory "is based on Ryan's (1970) premise that conscious goals affect action"
(Locke & Latham, 1990, 2002, p. 705). According to the theory, goals affect action in form of
choice, intensity, and persistence (Latham, 2007, p. 176). In addition, goals motivate individuals
to develop relevant strategies for goal attainment (Locke, 2001). Goal setting theory emerged
from empirical research over four decades, with more than 1,000 published articles and reviews
(Mitchell, Thompson, & George-Falvy, 2000) that have shown positive results in about 90% of
the studies (Locke, 2001). "The research is uniform in its verdict that difficult and specific goals
result in higher levels of performance than do easy or vague, 'do your best' goals" (Mitchell &
Daniels, 2003, p. 231). A model outlining the constructs associated with goal setting theory is
provided in Figure 1-1. Locke and Latham (1990; 2002) have provided thorough descriptions of
the theory and its related research. The theory is further described in chapter 2 of this
dissertation. Although the theory has been widely applied in for-profit organizations, it has been
given little consideration in the context of nonprofit organizations and volunteer-run groups.
Nonprofit Organizational Effectiveness
An important resource for tackling social issues, including those associated with
agricultural leadership, is nonprofit organizations. The missions of these organizations are
diverse, but their impact can be powerful. Although nonprofits represent only 7 percent of the
interest groups in Washington DC, from the 1960s through the 1990s nonprofit groups
"accounted for anywhere from 24 to 32 percent of the congressional testimony, generated
between 29 and 40 percent of the press coverage of pending legislation, and were nearly 80
percent as effective in passing legislation they favored as the business lobbies against which they
were often arrayed" (Salamon, 2002b, p. 43-44).
Herman and Renz conducted a panel study of nonprofit organizations (NPOs) in 1994 and
again in 2000. From the data collected, they have derived nine theses on organizational
1. Nonprofit organizational effectiveness is always a matter of comparison.
2. Nonprofit organizational effectiveness is multidimensional.
3. Boards of directors make a difference in the effectiveness of nonprofit organizations,
but how they do so is not clear.
4. Nonprofit organizational effectiveness is a social construction.
5. The more effective NPOs are more likely to use correct management practices.
6. Claims about "best practices" for nonprofit boards and for the management of NPOs
warrant critical investigation.
7. A measure of NPO effectiveness that emphasizes responsiveness may offer a solution
to the problem of differing judgments of effectiveness by different stakeholder groups.
8. It can be important to distinguish among different "types" of nonprofit organizations in
order to make progress in understanding the practices, tactics, and strategies that may
lead to NPO effectiveness.
9. Nonprofit organizations increasingly operate as part of networks of service delivery.
Therefore, network effectiveness is becoming as important to study as organizational
effectiveness. (Herman & Renz, 2002)
Although these theses offer improved understanding and implications for nonprofit
organizations in general, more research is needed to understand the characteristics of
effectiveness for nonprofit organizations and associations that address the need for agricultural
leadership. This is particularly important, since "the more general tendency of nonprofit scholars
is to ignore membership organizations and associations, especially grass-roots associations"
(Smith & Shen, 1996, p. 285). These voluntary associations are an important resource for
recruiting and developing local leaders (Bolton, 1991), yet there is no consensus on the best
approach for maximizing that leadership development process.
Farm Bureau's Leadership Needs
This study will address the field of agricultural leadership in the context of a specific
nonprofit agricultural leadership organization the Florida Farm Bureau Federation (FFBF).
The organization is unique in its grassroots approach to solving problems in the agricultural
industry and the communities that the organization serves. The grassroots approach is
characterized by the organization's policy development and implementation process, which is
driven by local members (Figure 1-2).
FFBF is part of a larger, national organization known as the American Farm Bureau
Federation (AFBF). In part, the mission of the AFBF is "to implement policies that are
developed by members" (American Farm Bureau Federation, 2006). Although the problem of
maintaining "grassroots" efforts is a challenge for the entire national organization, FFBF has
been proactive with statewide efforts to improve the flow of ideas from individual members to
the larger organization. This goal is made clear in FFBF's vision of being "the most effective,
influential and respected Farm Bureau in the nation" and to "be recognized as Florida's 'Voice
of Agriculture'" (FFBF, 2005b). FFBF has identified the leadership role of local officers in
county Farm Bureau groups as pivotal to organizational success (H. S. F. Carter, 2004).
A needs assessment conducted in 2003 revealed that FFBF state leaders believe four local
organizational aspects are important for an effective grassroots process: leadership, political
process, effective boards, and knowledge of Farm Bureau. When evaluating these areas with
local members, Carter (2004) applied a modified Borich (1980) needs assessment and found
significant differences between perceived importance and proficiency in the areas of leadership,
political process, and knowledge of Farm Bureau. For example, in the area of leadership, the
competency "use effective communication skills in media interviews" was rated on a seven-point
scale with a mean importance of 5.6 and a mean proficiency of 4.8, leaving a "gap" of 0.8. The
difference suggests a desire for improved proficiency with that competency. Within the area of
political process, participants rated "develop relationships with elected officials on the national
level" with a mean importance of 5.2 and a mean proficiency of 3.9, leaving a gap of 1.3
between importance and proficiency. Among the competencies related to knowledge of Farm
Bureau, participants rated "demonstrate knowledge of the American Farm Bureau Federation"
with a mean importance of 5.2 and a mean proficiency of 4.2, leaving a gap of 1.0. Carter
concluded that the findings indicate areas for leadership training with local FFBF board
members. In addition, Carter concluded, "this study could be the starting point for additional
leadership research within the FFBF" (p. 167).
In 2004, a qualitative study was conducted to further determine the leadership
expectations, needs, and interests of local FFBF board members. Local board members
throughout the state were interviewed, focusing on identification of common leadership-related
challenges and perceived development needs of the local Farm Bureau board. Four significant
theme areas emerged: organizational appreciation, grassroots involvement, board member
training, and board member succession. Many of the statements made by county board members
matched up with the expectations for board effectiveness previously identified by FFBF state
leaders (H. S. F. Carter, 2004). The theme areas described by Carter (2004) (leadership, political
process, effective boards, and knowledge of Farm Bureau) were well-represented in the
interviews with local board members. In addition, the findings from the interviews with local
board members seemed to further support the need for and interest in professional development
programming for FFBF's local leaders. Based on the findings, the researchers recommended
that the FFBF invest in the development of an educational program focused on FFBF's local
leaders (Kaufman & Rudd, 2006).
This recommendation for leadership development is consistent with interventions for
improved organizational effectiveness recommended for other non-profit and volunteer-based
organizations. For example, based on a three-year study of 24 diverse non-profit organizations,
Holland and Jackson (1998) concluded that ongoing efforts to develop board skills can markedly
improve their performance. Brudney and Murray (1998) add that "although these efforts may
not correct all problems (such as those caused by having the 'wrong' kinds of people on the
board), they can improve the chances of identifying, and then correcting, a number of others" (p.
346). Tierney (2006) argues that investment in leadership capacity is the "single most important
determinant of organizational success" (p. 3). Even still, a common conclusion among research
on interventions with volunteers and non-profit boards is that more empirical evidence is needed
to support and explain the connection between board-member and organizational effectiveness
(Brudney & Murray, 1998; Herman & Renz, 1999; Holland & Jackson, 1998).
"Strengthening the Voice" Program
In 2005, FFBF contracted with the University of Florida Department of Agricultural
Education and Communication (UF-AEC) to develop curricula for half-day workshops focused
on developing local Farm Bureau leaders. The specific focus was with local board members.
Later titled "Strengthening the Voice" (STV), the program was to include workshops that cover
the topics of effective meetings, political advocacy, member development, organization
management, and Farm Bureau knowledge. The finished curriculum was presented in a train-
the-trainer format, preparing FFBF staff to deliver the workshops locally throughout the state.
STV program developers and FFBF advisors determined that all of the workshops being
developed should use a "best practices" approach, thus highlighting proven techniques for
success that participants could adopt to improve upon their current levels of success. Nonprofit
scholars have cautioned against use of a "one best way" approach to management and board
practices for all organizations. Instead, "every organization must discover and continually seek
to improve its practices, consistent with its values, mission, and stakeholders' expectations
(Herman & Renz, 2004, p. 702). Accordingly, the best practices promoted in the STV program
were based on previous research with FFBF's state and local leaders (H. S. F. Carter, 2004;
Kaufman & Rudd, 2006) and were further refined with a proj ect advisory committee that
included FFBF staff and leaders.
Beyond the formal workshop experience, the STV program emphasized the importance of
application and follow-up. The general approach of the program was to (1) conduct a workshop,
(2) have participants set goals for adopting specific best practices, and (3) provide follow-up
reminders to participants regarding their best practices goals and other important points from the
program (Figure 1-3.) The incorporation of goal-setting is an adaptation of goal setting theory
(see Locke & Latham, 2002) and makes use of a finding by Holland and Jackson (1998) that the
most useful interventions with non-profit boards often link process with substance by asking
board members to set goals for themselves.
The STV program follows a logic model with expected short-, intermediate-, and long-
term outcomes. In the short-term, program planners expect that local board members will be
knowledgeable about best practices for board operations and involvement. Six to 12 months
after program participation, the expected intermediate outcome is that local FFBF boards will
have improved efficiency and productivity, which will in turn create increased grassroots
involvement in the FFBF. The long-term program goals are that (1) FFBF will be a nationally
recognized leader agricultural problem solving and policy formation and (2) FFBF members will
have increased community involvement that improves the social and economic structures in their
communities. These desired outcomes of the STV program aligned with the FFBF mission and
vision, as is shown in Figure 1-4.
As with any program, there are confounding factors to consider when implementing and
evaluating the success of the STV program. Specifically, the program planners were concerned
about factors that would limit local board members level of engagement in the identified best
practices. For some board members, unexpected or unusual changes and commitments in their
personal lives or professional careers may limit the time and energy they are able to devote to
implementation of the best practices. In addition, some board members may be hesitant to apply
the practices because they believe they do not have enough experience or expertise with the
recommended practices. This may be of particular concern considering that the program
workshops are designed to last only four hours, allowing for only a limited opportunity to apply
the practices in a trial setting. Another potential confounding factor relates to local board
members relationships with one another. Many of the identified best practices require
implementation at a group level in order for the desired program outcomes to occur. If local
board members do not trust one another and cooperate well together, they may have a difficult
time collaborating on the recommended practices.
"Farm Bureau Foundations" program component
The first STV component implemented was entitled "Farm Bureau Foundations." The
obj ectives of the workshop were:
1. Explain the relationship between Florida Farm Bureau Federation and Insurance;
2. Establish a local strategy for active pursuit of the Florida Farm Bureau Federation
3. Describe the influence and limitation of the Farm Bureau on political issues; and
4. Prepare a plan for grassroots Farm Bureau policy development and implementation.
The best practices promoted through the Farm Bureau Found'ations workshop were
developed from the needs assessment Eindings (H. S. F. Carter, 2004; Kaufman & Rudd, 2006)
and from discussion among the STV advisory committee members. The advisory committee
consisted of FFBF staff and UF-AEC representatives. The FFBF staff brought extensive
knowledge of the organization and experience with the local FFBF board members. The UF-
AEC representatives were familiar with Farm Bureau, but the primary value of their involvement
was with their expertise in writing obj ectives and establishing goals for leadership programming.
This approach to identification of best practices coincides with an approach promoted by
Herman and Renz (2004). They argue that "every organization must discover and continually
seek to improve its practices, consistent with its values, mission, and stakeholders' expectations"
(p. 702). The best practices identified for the FarmFFFFF~~~~~~~FFFFFF Bureau Found'ations program piece are
outlined in Table 1-1.
FFBF staff implemented the Farm Bureau Found'ations program component with twelve
different four-hour workshops throughout the state of Florida during April and May of 2006.
The program focused on local board members in FFBF geographic districts across Florida and
drew 156 participants from 45 different Florida counties. Although the FFBF staff members
promoted participation in the program by all members, the goal was to keep individual
presentations to a manageable size of about 15 to 20 participants, with about 25 to 30
representatives from each district. The dates of the workshops and number of participants in
each are outlined in Table 1-2.
The Farm Bureau Found'ations workshops concluded with participants setting personal
goals for best practices they planned to implement in the coming months. This "Best Practices
Goal Sheet" is displayed in Appendix A. These goal sheets were collected by the workshop
presenters and kept for a follow-up mailing. However, the number of goal sheets collected was
noticeably less than the number of identified workshop attendees, which suggests some degree of
program implementation failure. Unfortunately, this variation in program implementation has
the potential to present inconsistencies in program effects.
After workshop evaluation sheets indicate that over 98 percent of participants were
"satisfied" or "very satisfied" with the workshop. When asked to what extent they could use the
ideas and skills learned in the workshop, 59 percent answered "to a great extent" and 37 percent
answered "to a moderate extent." One participant wrote, "I've been wanting/needing this info
In June of 2006, the FFBF Hield staff returned each "Best Practices Goal Sheet" to the
participant who originally completed it. The cover letter of the mailings thanked the participants
for their service to FFBF, reminded them of the best practices commitments that they made, and
offered to provide assistance in accomplishing the best practices goals. A template for these
cover letters is displayed in Appendix B. In July of 2006, FFBF mailed program participants a
two-page newsletter that thanked the members for their participation in the program and
reminded them of the key points covered in obj ectives one and two of the Farm Bureau
Foundations workshop. In August of 2006, FFBF mailed a second newsletter that highlighted
the key points covered in objectives three and four of the Farm Bureau Found'ations workshop.
These newsletters are displayed in Appendix C. Although the intent was that all program
participants would receive these mailings, inconsistent records of attendance prevented some
participants from receiving the follow-up mailings.
The primary impetus for the STV program was a desire to improve the effectiveness of
local FFBF boards of directors. An important question then is whether or not aspects of the
program are influencing board effectiveness. Did any attitude and behavioral changes occur?
To a broader degree, though, can changes in board performance be attributed to constructs
associated with goal setting theory?
FFBF invested substantial time, money, and effort in the implementation of the Farm
Bureau Found'ations component of the STV program. Was it worth the investment? Are the
participants more proficient in FarmFFFFF~~~~~~~FFFFFF Bureau Found'ations related tasks? FFBF has plans to
quadruple this investment through implementation of four more components of the STV
program. Are there ways to better focus these efforts? What participant and program
characteristics are most associated with achievement of the best practices goals? Does the
answer lie in a more direct application of goal setting theory?
With an ex post facto study, it is not possible to directly measure change and answer all of
these questions. However, these inquiries lead to an important research question: What factors,
variables, or interventions explain the quality of engagement (in terms of best practices) of local
board members in a multi-level, membership organization? This question was used to guide the
exploratory study with the STV program.
The purpose of the study was to explain local board performance, as related to a
professional development program for local boards of a nonprofit organization.
The obj ectives of this study were as follows:
1. Describe select demographics of local Florida Farm Bureau board members and
2. Measure the engagement level of local Florida Farm Bureau board members in
FarmFFFFF~~~~~~~FFFFFF Bureau Found'ations Best Practices;
3. Measure the goal attributes experienced by program participants with their Farm
Bureau Found'ations Best Practices goal;
4. Explain individual-level performance with best practices identified in the FarmFFFFF~~~~~~~FFFFFF
Bureau Found'ations program; and
5. Explain board-level performance with best practices identified in the Farm Bureau
Significance of the Study
This study adds to the body of knowledge on goal setting theory, rural leadership
development, volunteer management, and effective board member practices. Practitioners in
these areas will use this study to guide future professional development programming efforts and
to guide further research.
Past research involving goal setting theory has focused primarily on its application in work
settings. If goal setting theory is shown to have application in volunteer settings, the theory can
be used to guide the training and management of volunteers and leaders of volunteers. Latham
and Kinne (1974) have shown that training employees to set specific, high goals results in
improved performance, but little has been done has been done to extend this research to
volunteer situations. The Eindings of this study may further that effort.
Regardless of the findings specific to goal setting theory, FFBF will make direct use of this
study's findings to make revisions to the STV program components in order to increase their
effectiveness. The improvement of the STV program will aid FFBF in accomplishing its mission
and vision, which in turn contributes to improved social and economic structure in the
communities served by FFBF.
In addition, plans are being made to adapt individual components of the STV program for
use in other states and organizations. The Eindings of this study will aid in the adaptation and
promotion of those program components for use by other groups. As other membership
organizations look for guidance in the selection and involvement of board members, they may
look to the Eindings of this study. Such organizations may be similar to Farm Bureau in their
structure, focus, or clientele. For example, local nonprofit organizations or agricultural
commodity associations or rural development groups may look to this study's discussion of
board member engagement when seeking to improve the involvement of their local board
As with any scholarly study, there are limitations to the generalizability of this study. The
first limitation that must be considered is related to the nature of the organization being studied.
Farm Bureau is a unique organization in terms of its structure and function. As a result, the
conclusions of this study have limited generalizability to other organizations. However, the
conclusions from this study may provide helpful information for practitioners interested in
conducting similar studies and programs with similar organizations.
Second, this study is limited to the study participants. The study design investigates
relationships among variables and does not indicate a cause-effect relationship. As a result, the
study's findings cannot be generalized to future program participants. Study participants were
self-selected into the program and may differ in important characteristics from non-participants.
Future evaluations must be conducted to ensure that similar relationships exist among future
The third limitation of the study deals with the limited number of variables included in the
study. Other demographic variables and program characteristics that were not included in this
study may influence goal achievement, and thus provide a different set of findings. The
variables included in this study were based on prominent social science theories.
The final limitation of the study is related to the validity and reliability of the self-reported
data examined in this study. A certain degree of error is associated with all survey studies, and
that error places limitations on the accuracy of the findings. These limitations are further
explained in the methodology and results sections of this study.
The following terms and definitions apply to this study:
* Agricultural Lead'ership The process of providing vision and advocacy for the business
activities associated with the production and marketing of plant and animal products.
* Andragogy A set of "core principles of adult learning that in turn enable those designing
and conducting adult learning to build more effective learning processes for adults"
(Knowles, Holton, & Swanson, 2005, p. 2)
* Farm Bureau Found'ations Best Practices The Farm Bureau member practices promoted
through the Florida Farm Bureau Federation's "Farm Bureau Foundations" workshop.
* Farm Bureau Found'ations program component The workshop and follow-up activities
associated with the first segment of Florida Farm Bureau Federation' s Strengthening the
* Farm Bureau Found'ations workshop A Farm Bureau knowledge-focused workshop
conducted by Florida Farm Bureau Federation staff during April or May of 2006.
* Florid'a Farm Bureau Fed'eration (FFBF) Florida' s largest general farm organization,
consisting of more than 140,000 grassroots member-families who are also members of the
American Farm Bureau Federation (FFBF, 2007b).
* Goal Conanitnzent An attribute of goal setting that is closely associated with goal
intensity and represents the degree to which an individual is determined to accomplish a
task (Locke & Latham, 1994).
* Goal Content A general attribute of goal setting that includes goal specificity and goal
difficulty (Locke & Latham, 1990).
* Goal Difficulty A measure of goal "level" that represents an individual's perception of
how challenging it is to accomplish a particular goal (Locke & Latham, 1990).
* GoalhIntensity "The scope, clarity, and mental effort involved in mental processes"
associated with goal setting and achievement, most often evaluated as goal commitment
(Locke & Latham, 1994, p. 16).
* Goal Specificity The degree to which goal content is vague (i.e. "work on this") or
concrete (i.e. "try for a score of 25 on this task within the next ten minutes") (Locke &
Latham, 1994, p. 15).
* Goal-Setting Theory A theory of adult motivation based on the premise that conscious
goals affect action (Locke & Latham, 2002).
* Grassroots Organization A nonprofit association or group that maintains a local
membership of volunteers who perform most of the work/activity done in and by the
nonprofit (Smith, 2000).
* Self-Efficacy Task-specif ic self-conf idence, which is based on a relationship between
perceived ability and task difficulty (Bandura, 1986).
* Sor Ie~llilthni the Voice (STV) program A long-term professional development program
conducted by the Florida Farm Bureau Federation for its local, county leaders.
This chapter provided the background and significance of the problem as well as the
purpose of the study. Agricultural leadership development was highlighted as an important tool
for securing a positive future for the agricultural industry and the communities where agriculture
is found. In addition, goal setting theory was introduced and recent research with the Florida
Farm Bureau Federation was highlighted. This led to a discussion of the FFBF .So Iengthen~ling the
Voice program and the Farm Bureau Found'ations program component, which served as the
context for the study.
The study's research question was defined as "What factors, variables, or interventions
explain the quality of engagement (in terms of best practices) of local board members in a multi-
level, membership organization?" The purpose of the study was to explain local board
performance, as related to a professional development program for local boards of a nonprofit
organization. The objectives of this study were as follows:
1. Describe select demographics of local Florida Farm Bureau board members and
2. Measure the engagement level of local Florida Farm Bureau board members in
FarmFFFFF~~~~~~~FFFFFF Bureau Found'ations Best Practices;
3. Measure the goal attributes experienced by program participants with their Farm
Bureau Found'ations Best Practices goal;
4. Explain individual-level performance with best practices identified in the FarmFFFFF~~~~~~~FFFFFF
Bureau Found'ations program; and
5. Explain board-level performance with best practices identified in the Farm Bureau
This chapter included the operational definitions of several terms related to the study.
Limitations of the study were also discussed. Chapter 2 will present the theoretical and
conceptual frameworks that guided this study.
Commit to New
(e.g., Performance and
Figure 1-1. Essential elements of goal-setting theory and the high-performance cycle. Note.
From "Building a Practically Useful Theory of Goal Setting and Task Motivation: A
35-Year Odyssey," by E. A. Locke and G. P. Latham, 2002, American Psychologist,
57(9), p. 714. Copyright 2002 by the American Psychological Association, Inc.
Reprinted with permission.
Policy Development andr Imnplementation Process
Implem autation a
5. Policies Adopted
Figure 1-2. Florida Farm Bureau Federation policy development and implementation process.
Note. From Florida Farm Bureau Federation's Agricultural Policy Division.
Reprinted with permission.
Program presentation teams consult
with local boards and schedule locally
Field staff recruit participants at local
board meetings and other Farm
practices adoption goal
SPresenters provide guidance for goal
Setting related to the best practices
Field staff provide informal feedback
to local board members
SFFBF staff evaluate and fine-tune the
Presentation teams deliver STV
0 Best practices promoted
0 Participatory activities
0 Reflection & discussion
0 Take-home manual
FFBF local leaders live
in coverage area
Local leaders receive
Local leaders participate
in workshop and receive
0 Needs assessment
0 Outline program
0 Train presenters
0 Assist in
State staff distribute follow-u]
0 Personal goal sheet
Figure 1-3. Process model for the Florida Farm Bureau Federation "Strengthening the Voice" program.
0 Allocate staff
0 Allocate resources
0 Monitor program
Workshops and Specific
Activities with I IComponent
Local FFBF Objectives
FL Farm Bureau Mission:
Joint and collective effort at the county,
state, and national level to solution of
problems of agriculturalists and rural
FL Farm Bureau Vision:
4 Voice & Leader of FL Ag
4 Well informed, educated, and
4 Increased net income to members
+bb~ Is a Nationally Kecognizec
Leader in Agricultural Problem
Solvina and Policy Formation
FFBF Local Leaders
Knowledgeable of and Improved Efficiency &
'Engaged in Best Practices Productivity of FFBF
Ifor Local FB Operations & Local Boards
Figure 1-4. Florida Farm Bureau (FFB) "Strengthening the Voice" program logic model.
Table 1-1. FarmFFFFF~~~~~~~FFFFFF Bureau Foundations workshop best practices by objective
Obj ective # Best Practice
1 Use Farm Bureau facts and history in member recruitment.
Share Farm Bureau facts and history in public relations activities.
Recognize and promote member benefits (like insurance).
Limit board meeting discussion to federation business.
Allow time on meeting agendas for an insurance report.
Maintain a working relationship with the insurance agency manager.
2 *Actively use Farm Bureau's grassroots approach to governance.
Cite Farm Bureau's mission when promoting the organization to potential
Hold Farm Bureau leadership accountable to the organization mission, vision,
Adopt a local mission and values to guide board actions.
3 *Promote continued political involvement by citing Farm Bureau's past
Maintain organizational accountability to the grassroots mission when new
policies are proposed.
Maximize the benefits of collaboration when promoting political policy.
4 *Implement grassroots policy by following Farm Bureau's development and
Strengthen the "Voice of Agriculture" by getting involved in statewide
Plan for policy implementation challenges before they arise.
Consult others, including adversaries, for input and feedback on policy ideas.
Table 1-2. FFBF FarmFFFFF~~~~~~~FFFFFF Bureau Foundations workshop dates and participant numbers by district
District Workshop Date # of Participants*
1 May 25, 2006 19
1 &2 April 06, 2006 15
2 April 13, 2006 22
3 May 30, 2006 10
4 May 11, 2006 7
4 &5 May 18, 2006 18
6 April 11, 2006 20
6 April 20, 2006 7
7 May 02, 2006 8
7 May 11, 2006 8
6 &8 May 19, 2006 13
8 May 11, 2006 9
Total 12 workshops 156
Some participants were not local board members, so the number of participants includes over-
REVIEW OF LITERATURE
Books and research-based articles on professional development include a wide range of
theories and practices. Of these approaches, theories of motivation seem to be at the core.
Furthermore, literature reviews reveal that the principle of goal setting is the one overriding
common theme recognized, either implicitly or explicitly, among virtually all psychological
approaches to motivation (Austin & Vancouver, 1996; Latham, 2007; Locke & Latham, 1984;
Mitchell & Daniels, 2003). Professional development is essentially about helping individuals
and organizations achieve their goals (J. E. Osborne, 2003). Not surprisingly, goal-setting theory
has developed into one of the most practical and empirically sound theories of motivation
(Latham, 2007; Locke & Latham, 2002; Miner, 1984; Pinder, 1998).
Goal Setting Theory Defined
Goal setting theory was formulated based on empirical research over four decades. "It is
based on Ryan's (1970) premise that conscious goals affect action" (Locke & Latham, 2002, p.
705). Within the context of the theory, "a goal is the obj ect or aim of an action" (Locke &
Latham, p. 705). According to Latham (2007), "goals are the immediate precursor of action" (p.
176). The theory applies specifically to performance and learning goals that involve degrees of
success with particular tasks. Goals affect action in form of choice, intensity, and persistence
(Latham, p. 176). In addition, goals motivate individuals to develop relevant strategies for goal
attainment. Locke and colleagues summarized the early research on goal setting theory as
Goal setting is most likely to improve task performance when the goals are specific and
sufficiently challenging, the subjects have sufficient ability (and ability differences are
controlled), feedback is provided to show progress in relation to the goal, rewards such as
money are given for goal attainment, the experimenter or manager is supportive, and
assigned goals are accepted by the individual. (Locke, Shaw, Saari, & Latham, 1981, p.
These findings continued to be replicated throughout research over 35 years (Locke & Latham,
2002). "Goal setting studies have been conducted with 88 different tasks including bargaining,
driving, faculty research, health promoting behaviors, logging, technical work, managerial work,
management training, safety, and sports performance" (Locke & Latham, 1994, p. 16). Goals
have been found to increase performance at individual, group, and organizational levels
(Mitchell & Daniels, 2003).
Goal setting and planned behavior
Ajzen's theory ofplanneed behavior (Ajzen, 1985, 1991; Ajzen & Madden, 1986; Schifter
& Ajzen, 1985) provides a helpful context for considering goal setting theory. According to the
theory of planned behavior, "the more favorable the attitude and subj ective norm with respect to
a behavior, and the greater the perceived behavioral control, the stronger should be the
individual's intention to perform a behavior under consideration" (Ajzen, 1988, p. 132-133). A
structural model of the theory of planned behavior is shown in Figure 2-1.
In the context of the theory of planned behavior, the "intention" is a self-selected goal.
When a goal is self-selected, goal setting theory can be used to more fully explain the connection
between intention and behavior. In fact, goal setting theory is based on Ryan's (1970) premise
that behavior is regulated by intentions (Latham, 2007). However, goal setting theory presents a
more comprehensive approach and more detailed explanation of this relationship than does the
theory of planned behavior. Beyond self-set goals, goal setting theory can be used to explain
assigned goals that may be the precursor to the components of the theory of planned behavior.
According to goal setting theory, goals affect performance through four mechanisms: (1)
choice/direction, (2) effort, (3) persistence, and (4) strategies (Locke & Latham, 2002). "Goals
direct attention and effort toward goal-relevant activities and away from goal-irrelevant
activities." In addition, "goals have an energizing function. High goals lead to greater effort than
low goals" (p. 706). Time on task is related to effort and persistence. "When participants are
allowed to control the time they spend on a task, hard goals prolong effort" (p. 707). "Goals
affect action indirectly by leading to the arousal, discovery, and/or use of task-relevant
knowledge and strategies" (p. 707).
One of the most extensively studied attributes of goals is goal content. Research studies
involving goal setting theory have consistently found that specific, difficult goals consistently
lead to high performance (Latham, 2007; Locke & Latham, 1990, 2002; Locke et al., 1981;
Mitchell & Daniels, 2003). Locke and Latham provide some clarification on these aspects of
goal content: specificity and difficulty.
With regard to specificity, goal content can be vague ("work on this") or specific ("try for a
score of 25 on this task within the next ten minutes"). With regard to goal difficulty, goals
can be easy ("try to get 10 items completed in the next 20 minutes"), moderate ("try to get
20 ."), difficult ("try to get 30 ."), or impossible ("try to get 100 ."). Difficulty
pertains to a relationship between the person and the goal. The same goal can be easy for
one person and hard for another depending on ability and experience. Generally, however,
the higher the absolute level of the goal, the more difficult it is for people to attain it.
(Locke & Latham, 1994, p. 15)
Locke and Latham (1990; 1994; 2002) report a positive, linear relationship between goal
difficulty and levels of performance. The explanation for this finding is that "people adjust their
effort to the difficulty of the task undertaken" (Locke & Latham, 1994, p. 15). "Goal theory
predicts and finds a performance drop at high difficulty levels only if there is a large decrease in
goal commitment (or a poor task strategy was used)" (Locke & Latham, 1994, p. 15)
Goal specificity, in itself, does not lead to high performance, because specific goals can
still vary in difficulty. However, "goal specificity does reduce variation in performance by
reducing the ambiguity about what is to be attained" (Locke & Latham, 2002, p. 706). Goal
specificity and difficulty are the primary components of goals highlighted within goal setting
theory. Research in field settings supports the idea that people can be taught to set specific, high
goals (Latham & Kinne, 1974).
Goals have been further identified as learning or performance goals, yet the evidence to
support one goal type over another tends to be situation dependent (Latham, 2007; Seijts &
Latham, 2001; Seijts, Latham, Tasa, & Latham, 2004). "Goal setting without adequate
knowledge is useless" (Latham, 2007, p. 176). Therefore, in complex situations where
individuals lack preexisting knowledge of appropriate task strategies, setting a specific high
learning goal leads to higher performance than setting a specific high performance goal or a
vague "do your best" goal (Seijts et al., 2004).
Some research on goal setting theory has investigated the effects of distal versus proximal
goals. Bandura (1997) noted that distal goals can be too far removed to effectively guide present
action. Studies have shown, however, that performance improves when proximal goals are set in
addition to a distal outcome goal (Bandura, 1986; Latham & Seijts, 1999; Seijts & Latham,
Another extensively studied attribute of goal is goal intensity. "Intensity refers to the
scope, clarity, and mental effort involved in mental processes" (Locke & Latham, 1994, p. 16).
Goal intensity is most often characterized and studied as goal commitment, yet a related
construct worth measuring is self-efficacy. According to Locke and Latham, "commitment
refers to the degree to which an individual is attracted to the goal, considers it important, is
determined to attain it, and sticks with it in the face of obstacles" (p. 16).
Commitment is enhanced when people believe that achieving the goal is possible, and that
achieving the goal is important (Klein, 1991). The first class of factors raise what Bandura
(1986) has labeled self-efficacy (task-specific self-confidence). These include ability,
experience, training, information about appropriate task strategies, past success, and
internal attributions (Locke & Latham, 1990). Those in authority can affect the second
class of factors by explaining why the goal is important, exerting reasonable pressure for
performance, being knowledgeable about the task and job, and serving as a role model for
the behavior they desire in the subordinate (Locke & Latham, 1990). (Locke & Latham,
1994, p. 17)
Beyond these insights on goal intensity, Locke and Latham (2002) provide three
suggestions for improving self-efficacy: "(a) by ensuring adequate training that provides success
experiences, (b) by role modeling or finding models with whom the person can identify, and (c)
through persuasive communication that expresses confidence that the person can attain the goal"
(Locke & Latham, 2002, p. 708). Consistent with social cognitive theory (Bandura, 1986) and
the theory of planned behavior (Ajzen, 2005), perceived ability, rather than actual ability, is the
crucial motivating factor (Latham, 2007). For a more thorough discussion of self-efficacy, see
Beyond the goal intensity attribute, expressed by goal commitment and self-efficacy, goal
setting theory suggests that goal performance is moderated by other factors. These include
feedback, task complexity, personal ability, and situational constraints (Locke & Latham, 1994).
Of these, feedback and task complexity are the most commonly considered (Locke & Latham,
Summary feedback is a moderator of goals in that the combination of goals plus feedback
is more effective than goals alone" (Locke & Latham, 2002, p. 708). When people do not know
their level of progress toward a goal, "it is difficult or impossible for them to adjust the level or
direction of their effort or to adjust their performance strategies to match what the goal requires"
(p. 708). According to Bandura and Ceryone (2000), "simply adopting goals, whether easy or
personally challenging ones, without knowing how one is doing seems to have no appreciable
motivational effects" (p. 203). However, feedback need not be formal; it can be internal or
external (Latham, 2007). Thus, self-evaluation can be an important moderating factor of goal
performance (Bandura & Ceryone, 2000).
As the complexity of goal-related tasks increase, goal performance becomes dependent on
the ability to discover appropriate task strategies. "Because people vary greatly in their ability to
do this, the effect size for goal setting is smaller on complex tasks than on simple tasks" (Locke
& Latham, 2002, p. 709). In situations where a variety of task strategies are applied toward goal
achievement, goal-strategy interactions are often found, with goal performance closely
associated with the use of effective strategies.
Numerous goal-setting studies have examined the influence of demographic variables
related to goal performance, but the Eindings have not provided the reliability necessary for
inclusion in Locke and Latham' s (2002) model for goal setting theory. Renn and Fedor (2001)
examined age, education, tenure, and gender but found that none of these variables were related
to goal performance.
However, Cullen and others (2004) studied goal-setting with children and found that
gender, ethnicity, socio-economic status (SES), and mother' s education were influential
variables. Male students were found to have more success with general goal setting than female
students. Goal performance differences associated with ethnicity, SES, and education varied
somewhat with the specific goal task, but these differences were also closely aligned with
baseline preferences and activities. The researchers did recognize the uniqueness of their
findings and noted the need for further research in this area (Cullen et al., 2004).
Jeffries and Hunte (2004) suggest that generational differences act as an influence on work
motivation, which in turn impacts goal-setting and performance. For example, the "boomer"
generation (born roughly between 1946 and 1964) tends to associate time spent on a task with
successful performance, while generation "Xers" (born roughly between 1961 and 1981) tend to
prefer autonomy and place more emphasis on results (O'Bannon, 2001). As a result of these
differences, longer timelines with proximal goals may be appropriate for boomers, whereas Xers
may perform better with opportunities for choice and shorter completion times (Jeffries &
Hunte). This research supports the inclusion of age as a demographic variable in goal-setting
Doest, Maes, Gebhardt, and Koeleijn (2006) studied personal goals and found evidence
that education level and age both impact goal outcomes. Specifically, more highly educated
respondents tended to report more emotional exhaustion, which in turn was negatively related to
attainment of personal goals. The same was true for younger respondents. However, the
demographics were presented as dichotomous variables, and the researchers recognized that as a
limitation of the study. They suggest that "future research would benefit from the use of
longitudinal designs, more control variables, and objectively measured outcomes" (p. 214).
Based on published findings and the recommendations of other researchers, demographic
variables appropriate for consideration in goal-setting studies include: education, ethnicity, job
tenure, age, and gender.
Andragogy and Adult Learning
Adult learning dates back to ancient times, when teachers like Confucius, Socrates, and
Cicero focused efforts on teaching adults. However, "for many years, the adult learner was a
neglected species" (Knowles et al., 2005, p. 35). In the early 1970s, "the concept that adults and
children leamn differently was first introduced in the United States by Malcom Knowles," and
along with that concept came the term "andragogy" (Knowles et al., p. 1). Andragogy has been
interpreted in many ways, but Knowles and his colleagues posit that "andragogy presents core
principles of adult learning that in turn enable those designing and conducting adult learning to
build more effective learning processes for adults" (p. 2). Andragogy is the "single most popular
idea in the education and training of adults" (Brookfield, 1986). "The six principles of
andragogy are (1) the learner' s need to know, (2) self-concept of the learner, (3) prior experience
of the learner, (4) readiness to leamn, (5) orientation to learning, and (6) motivation to leamn"
(Knowles et al., p. 3). These principles are not necessarily in contrast to pedagogy, literally
defined as "the art and science of teaching children" (Knowles et al., p. 36), instead the
principles of andragogy provide guidance in ways to capitalize on the experience and perspective
that adults bring to the learning situation. As individuals mature, their need to be self-directing
and use their experience in learning increases. Andragogy accounts for this difference and
provides an opportunity for this difference to be an asset in the learning process, rather than a
Knowles and colleagues (2005) suggest that "andragogy works best in practice when it is
adapted to fit the uniqueness of the learners and the learning situation" (p. 3). They divide the
goals for adult learning into three general categories: individual, institutional, and societal
growth. The traditional view of adult learning has been to focus on individual growth. For
example, some adult learning programs have sought to advance the employability skills of
individual adults (Birkenholz, 1999). However, adult learning is helpful for achieving
organizational and community goals as well (Rohs & Langone, 1993). The American Society
for Training and Development (ASTD) is a professional organization that focuses on linking
adult learning with organizational performance. In addition, many Extension programs use adult
learning as a tool for achieving community goals. One example is the training of adult
volunteers to serve youth involve in 4-H programming.
Motivation to Learn
Motivation is a concept that explains why people think and behave as they do. "Seeing
human motivation as purposeful allows us to create a knowledge base about effective ways to
help adults begin learning, make choices and give direction to their learning, sustain learning,
and complete learning" (W10dkowski, 1999, p. 2). When learners are motivated, the learning
process goes more smoothly, anxiety decreases, and learning becomes more apparent (p. 5).
W10dkowski suggests that adult motivation to leamn is the sum of four factors: success, volition
(i.e. choice), value, and enjoyment (p. 14).
W10dkowski's factors for adult motivation are reflected in two of the principles of
andragogy. The first principle states that adults need to know why they need to learn \Ilm~inehig
before learning it, and the sixth principle is that the motivation for adult learners is internal
rather than external (Knowles et al., 2005). "Adults are responsive to some external motivators
(better j obs, promotions, higher salaries, and the like), but the most potent motivators are internal
pressures (the desire for increased j ob satisfaction, self-esteem, quality of life, and the like)"
(Knowles et al., 2005, p. 68).
Expectancy theory suggests three aspects of individual motivation that also support
W10dkowski's assessment of key motivational factors. Described in learning terms, expectancy
theory proposes that "adult learners will be most motivated when they believe they can learn new
material (expectancy) and that the learning will help them with a problem or issue
(instrumentality) that is important to their life (valence)" (Knowles et al., 2005, p. 201).
In order to maximize learner motivation, W10dkowski (1999) proposes five characteristics
and skills to be demonstrated by adult educators: expertise, empathy, enthusiasm, clarity, and
cultural responsiveness (p. 26). He has conceptualized these skills in the Motivational
Framework for Culturally Responsive Teaching: establishing inclusion, developing attitude,
enhancing meaning, and engendering competence (p. 81). W10dkowski proposes 60
motivational strategies that connect to the Motivational Framework for Culturally Responsive
Teaching (see Table 8.1 in W10dkowski, 1999, p. 294-297). Using the strategies proposed by
W10dkowski (1999), conceptualizing a program plan that motivates learners becomes a rather
Many of the concepts for improving adult motivation to learn are reflected in the FFBF
Strlingthel~litii the Voice program. For example, the workshops for each program component
begin by taking 10 to 15 minutes to develop (or reinforce) a felt need to learn among the learners
and to engender confidence in the program. This is accomplished indirectly through the
previously established credibility of the presenters (FFBF staff members) and more directly
through group discussion of the learning objectives. In addition to personal examples that might
vary from one presentation to another, consistent program examples are provided in the form of
video segments in which FFBF members and leaders discuss the practical value of the topics
about to be addressed in the program. Throughout the program, learner motivation is reinforced
by involving the learners in activities and discussion. One of the goals of the first program
components, FarmFFFFF~~~~~~~FFFFFF Bureau Foundations, was to provide a valuable and enjoyable learning
experience that would motivate participants to identify future program components as a priority.
Program Planning and Evaluation
Critics of professional development programs argue that little thought is put into program
outcomes (Hustedde & Woodward, 1996). For this reason, many professional development
models have been created for use with adult programs. Two specific models provide insight into
program characteristics that were integral to the program being studied. The models are the
Andragogy in Practice Model (Knowles et al., 2005) and the Conceptual Programming Model
(Boone, Safrit, & Jones, 2002). While these models provide guidance for program planning, a
key part of program improvement is evaluation (Umble, 2007).
Andragogy in Practice Model
The andragogical model has been identified as a prototypical model, because of its broad
influence on the practice of adult learning (Cookson, Knowles, Nadler, & Nadler, 1998).
Foundational to the model is the idea that adult educators should not be teachers, but instead
facilitators of learning (Cookson et al., 1998; Knowles et al., 2005). In this way, learning is not
characterized by one-way communication. Instead, learners are involved throughout the entire
process. According to Knowles and colleagues, the process of andragogy involves eight
elements: (1) preparing learners, (2) climate, (3) planning, (4) diagnosing of needs, (5) setting of
objectives, (6) designing learning plans, (7) learning activities, and (8) evaluation (p. 116).
The Andragogy in Practice Model was developed in 1998 "as an enhanced conceptual
framework to more systematically apply andragogy across multiple domains of adult learning
practice" (Knowles et al., 2005, p. 148). There are three dimensions to the andragogy in practice
model: (1) goals and purposes for learning, (2) individual and situation differences, and (3)
andragogy: core adult learning principles (Knowles et al., p. 148). Goals for adult learning are
conceptualized as an outside ring on the model and are identified as goals for individual,
institutional, or societal growth. The next dimension of the model is displayed as a middle ring,
highlighting subject-matter differences, situational differences, and individual learner
differences. The core of the model focuses on six andragogical principles: (1) learners need to
know, (2) self-concept of the learner, (3) prior experience of the learner, (4) readiness to learn,
(5) orientation to learning, (6) motivation to learn (Knowles et al., p. 149).
The Andragogy in Practice Model has the benefit of application to a vast number of adult
learning situations. Practitioners can begin with learning goals for the situation and follow the
model inward, or they can begin with the andragogical principles and move outward to the
specific learning goals (Figure 2-2). Knowles himself has recognized that the andragogical
model is merely a helpful starting point for adult learning and does not constitute a complete
theory (Knowles et al., 2005, p. 163). Even still, andragogy does provide "an enduring model for
understanding certain aspects of adult learning" (Merriam & Caffarella, 1999, p. 278).
As the curriculum materials for the FFBF's S~imigtheninilg~ the Voice program were being
developed, components of the andragogy in practice model were considered and applied. The
outside dimension of the model, "goals and purposes for learning," was incorporated by ensuring
that the program was based on the needs assessments that were conducted prior to program
development. Throughout the curriculum development process, the goals and purposes for each
program component were continually referenced by the curriculum writers and reaffirmed by the
program advisory committee (consisting of FFBF leaders). The second dimension of the
andragogy in practice model, "individual and situation differences," was incorporated by
designing the Stl eigthell~itiig the Voice program to be locally delivered and adaptable to local
needs. The FFBF Hield staff members were most knowledgeable of the individual and situational
differences of the target population, so the program was designed in a way that would allow
them to deliver the program locally in a way that fit the local needs. The final dimension of the
andragogy in practice model, "core adult learning principles," was considered throughout the
process. The first three principles, (1) learners need to know, (2) self-concept of the learner, and
(3) prior experience of the learner, were particularly important in the planning process to ensure
that the program would meet its target audience "where they were." The three additional
principles, (4) readiness to learn, (5) orientation to learning, and (6) motivation to learn, were
built into the program through activities that involved action and reflection. Although the
andragogy in practice model was not always referenced directly, its components shine through in
the program materials and design.
Conceptual Programming Model
The Conceptual Programming Model, as developed by Boone and others (2002), has an
accompanying book to provide explanation. "The model includes three interconnected and
related subprocesses: (1) planning, (2) design and implementation, and (3) evaluation and
accountability" (Boone et al., 2002, p. 42). Each subprocess is further divided into two
dimensions. The planning dimensions are: (la) the organization and its renewal process and (lb)
linking the organization to its publics. Dimensions of design and implementation include: (2a)
designing the planned program and (2b) implementing the planned program. And the evaluation
and accountability dimensions are: (3a) evaluation and (3b) accountability. Within each of these
dimensions, the model promotes specific tasks that are described in significant detail in the book:
"Developing Programs in Adult Education: A Conceptual Programming Model" (Boone et al.,
This professional development model is successful in capturing many of the important
elements of other program development models. Although Boone's early model has been
criticized for having few direct tests of the model (see Long, 1998, p. 81), Boone' s most recent
book cites numerous publications, pilot demonstrations, and research studies that involved
application of the Conceptual Programming Model (see Appendix of Boone et al., 2002).
Within the context of the Florida Farm Bureau' s S~imisthel~litiig the Voice program, the
Conceptual Programming Model (Boone et al., 2002) was not strictly applied, but the stages of
planning fit none-the-less. The planning processes included a needs assessment that involved
organizational leaders and local stakeholders. The planning sub-process was then followed by
design and implementation that was directed by professionals from the University of Florida and
the Florida Farm Bureau Federation. The program is still in the process of being implemented,
but initial efforts for evaluation are currently being conducted, with the idea that the evaluation
findings will guide program improvement and offer some accountability to stakeholders.
Professional development programs can provide benefits at individual, organizational, and
society levels. Evaluating their impacts falls under the realm of social science, which recognizes
the complexities that limit scientific certainty about any question involving human behavior
(Ary, Jacobs, & Razavieh, 2002, p. 14). Even still, professional development programs can and
should be assessed in such a way that suggests some degree of certainty about their true impact.
"Contemporary concern over the allocation of scarce resources makes it more essential than ever
to evaluate the effectiveness of social interventions" (Rossi, Lipsey, & Freeman, 2004, p. 28).
Fortunately, many research studies have documented benefits of professional development
programs. Earnest (1996) used the Leadership Practices Inventory as a pre- and post-test for
participation in a community leadership program. His findings demonstrated that participants (a)
were more willing to take risks; (b) broadened their perspective of leadership
roles/responsibilities; (c) developed a greater appreciation for teamwork and collaboration; and
(d) learned to adapt their leadership styles. Sugrue and Rivera (2005) documented one
organization that doubled customer satisfaction in less than a year as a result of implementing a
training program in customer service. Another organization increased retention of new sales
people from 60 percent to over 90 percent after implementing a professional development
program for new hires (Sugrue & Rivera, p. 18).
Galloway (2005) has identified Kirkpatrick' s (1998) Four-Level Model as a dominant
approach to evaluating professional development programs. According to Kirkpatrick (1998),
the sequence of evaluating training programs involve four levels, each impacting the next, and
each being more difficult and time-consuming, yet more valuable (p. 19). The levels are: (1)
reaction, or how well learners were satisfied with the program; (2) learning, including participant
changes in attitude, improved knowledge, or increased skill; (3) behavior, or the changes in
behavior that resulted from the program; and (4) results, including impacts such as increased
production, improved quality, higher profits, decreased turnover, etc. Kirkpatrick' s model
provides a valuable framework for program evaluation and has repeatedly been proven useful
since its inception in 1959.
In the realm of goal setting theory, evaluation is a critical component. Level one
evaluation, reactions, is important for understanding several goal moderators, specifically goal
commitment, goal importance, and self-efficacy. Level two evaluation, learning, corresponds
with measures of the remaining two goal moderators: feedback and task complexity. Level three
evaluation, behavior, provides an assessment of all the goal mechanisms: choice/direction, effort,
persistence, and strategies. Finally, level four evaluation, results, highlights the purpose of goal
setting, improved performance. Of course, all of these components of goal setting theory can
occur without evaluation, but the only way to fully understand them is through an evaluation that
covers Kirkpatrick's four levels: reactions, learning, behavior, and results.
Return on investment
Although the Kirkpatrick method has been viewed as a comprehensive model for
evaluating impact, training professionals have often been called upon to take evaluation a step
further and calculate a return on investment (ROI) (Galloway, 2005; Phillips & Phillips, 2007).
In response to this need, Phillips (2005) has added a fifth level to Kirkpatrick' s model which
focuses on the assessment of the monetary value results yield, compared to the cost of training.
Phillips (2005) suggests that there are hundreds of studies that demonstrate ROI for
organizational investment in professional development. One example of ROI level impact is a
literary skills training program conducted by Magnavox. The program cost $38,233 to conduct,
but the benefits were valued at $321,600, leaving a net benefit of $283,367 a 741% ROI
(Galloway, 2005). Research conducted by Bassi and McMurrer (cited in Phillips & Phillips,
2005) further supports ROI for professional development programming. They found that firms
that made unusually large investments in employee education and training outperformed the S&P
500 by a factor of two (113 percent versus 55 percent) in the years following investment (p. 15).
Sugrue and Rivera (2005) documented an organization that implementing a training program for
its sales force and in turn experienced "a $375 million increase in revenues in a space where
sales had been forecasted to be flat" (p. 18). Rohs (2004) evaluated the Southern Extension
Leadership Development (SELD) and reported that every one-dollar spent in the SELD program
returned $2.86 (286%) in net benefits (p. 27).
Professional development programs can demonstrate ROI; however, as a fifth level of
evaluation, it is even more difficult and time consuming than the four previous levels identified
by Kirkpatrick (1998). Phillips (2005) estimates that the cost of evaluating a program all the
way through ROI "may represent as much as 5 percent to 10 percent of the entire project." In
conducting ROI level evaluation, Rohs (2004) provides a practical example that may serve as a
guide. ROI calculations are based on the model presented by Phillips (1992), and highlight six
steps: (1) collect data, (2) isolate the effects of training, (3) convert data to monetary values, (4)
identify intangible benefits, (5) tabulate program costs, and (6) calculate the ROI (Rohs, 2004).
Leadership Development in Context
Applications of leadership development occur in different contexts and contextual factors
sometimes influence program goals and design (Hannum, Martineau, & Reinelt, 2007). Three
inter-related program contexts include rural, agricultural, and grassroots leadership development.
Each of these program areas might focus on a different target audience, but they also overlap
considerably and all tie into the interests of a group like Farm Bureau. Even still, it may be
helpful to consider the literature contributions of each area separately.
Rural Leadership Development
In order to ensure success facing the challenges of adult education in rural leadership
development, quality research must be performed and utilized, but that research has been
lacking. A variety of publications exists on the topic of rural community development, and
many of the publications use the phrase "leadership development." However, little research has
focused on programmatic efforts for the development of rural leaders. In fact, an extensive
review of literature in Cambridge Scientific Abstracts yielded only 15 relevant research articles
over a decade (1994-2004) of publication (Kaufman & Rudd, in press). Among the articles
found, sub-topic areas identified included Partnerships, Political Shifts, Safety & Health, Gender
Equity, Public Discourse, Statewide Program Impact, and Continuing Education. However, each
area lacked adequate saturation of research. The effectiveness of leadership development in rural
areas is severely threatened by this lack of published research. With increasingly limited
resources, those pursuing rural leadership development must address significant deficiencies in
the research and share findings for the betterment of all (Kaufman & Rudd, in press).
For practitioners in rural leadership development, a helpful guide has been published by
Hustedde and Woodward (1996) through the University of Kentucky Cooperative Extension
Service. According to the report, "many social scientists, business leaders, and others recognize
that the key to addressing rural problems is the 'capacity building' of local leaders and citizens"
(.1.The model proposed for capacity building starts with the development of clear goals
around which activities can be structured. In addition, Hustedde and Woodward suggest basing
leadership programs on the concept of "community trusteeship." Community trustees are
servant-leaders who make "an active and caring commitment to the community" (p. 2). Essential
public skills to be taught in a rural leadership development program might include: active
listening, collaboration, conflict resolution, deliberation, evaluation, facilitation, imagination,
interviewing, negotiation, power analysis, strategic planning, team building, vigilance, visioning,
and volunteer management. Hustedde and Woodward caution against lecture formats in rural
leadership development programs and instead advocate for activities that stimulate critical
Dhanakumar, Rossing, and Campbell (1996) also advocate for activities that involve
critical thinking, and have observed concrete benefits from these activities. In their evaluation of
the Wisconsin Rural Leaders Program, they found that "rural leaders learn best by a process of
action and reflection" (p. 6). Previous studies on civic participation and socio-economic status
suggest that rural citizens, as a lower socio-economic class, might not engage in civic activities.
However, the research by Dhanakumar, Rossing, and Campbell found that the Wisconsin Rural
Leaders Program improved civic engagement among rural citizens. This finding provides
confirmation for previous research by Martin and Wilkinson (1985) who concluded that
leadership development mediates the relationship between public affairs participation and
socioeconomic background. Accordingly, "leadership development programs can have an
important redistributive effect," providing real economic benefits to rural citizens and
communities (Martin & Wilkinson, p. 106)
Agricultural Leadership Development
Agricultural leadership can encompass a broad reach of topics and programs. In reviewing
research with keywords of "agricultural" and "leadership," we Eind that subj ect areas include
leadership programs for rural youth, leadership education programs in colleges of agriculture,
adult agricultural leadership programs, governmental agricultural agency positions, and general
rural community development efforts. Among these topics, all require additional research to
further the Hield and scholarship of agricultural leadership (E. Osborne, 2007; Rudd et al., 2004).
Fortunately, the Hield has found a home and increased acceptance in the discipline of agricultural
education. More and more, research conferences and journals related to agricultural education
are welcoming research in the area of agricultural leadership. In addition, collegiate departments
of agricultural education are increasingly offering courses and programs in agricultural
leadership (F. W. Brown & Fritz, 1994; S. Fritz, Hoover, Weeks, Townsend, & Carter, 2003; S.
Fritz, Townsend et al., 2003; S. M. Fritz & Brown, 1998).
Although many agricultural leadership efforts are taking place throughout the world, the
research Eindings related to these efforts are too often unpublished. Fortunately, individuals
engaged in the Hield of agricultural leadership development are a collegial group and often share
unpublished research Eindings. For example, the Arkansas LeadAR program recently conducted
a full review of their efforts, summarizing all of their program materials and conducting focus
group interviews with program graduates and stakeholders (Black, Day, Roberts, Roseleip, &
Singletary, 2005). Although these research findings have not been formally published, they were
shared informally at the 2005 International Association of Programs for Agricultural Leaders
(IAPAL) conference. IAPAL members were invited to request additional details from the
research, and the LeadAR program director mailed full research findings to individuals who
Another example of adult agricultural leadership program research being shared informally
is recent research evaluating the Ohio LEAD program. Black, the Ohio LEAD program director,
informally shared some of her research design at the 2005 IAPAL conference. She has since
completed her evaluation and provided an executive summary of the findings to select IAPAL
members (Black, 2006). Black observed that 86% of participants recognized personal changes
as a result of program participation. Those changes were observed at three different levels:
individual, organizational, and community. Respondents reported being most affected by the
program at the individual level. The most highly reported change was in networking, with 63%
of respondents reporting an increase in their use of social networks. The benefits of this
increased networking carried over into their business activities, thus providing change at the
organizational level. In addition, respondents indicated that the program improved their
decision-making skills and their ability to facilitate change. At the community level, 75% of
respondents reported an increased appreciation for cultural diversity and 45% of respondents
reported increased involvement in local proj ects and/or organizations. Black and Earnest (2006)
presented some of the findings at the annual conference of the International Leadership
Association. As of February 2007, the findings of the study had not been published in any
research j ournals that might allow the information to be more widely used.
One evaluation of the agricultural leadership program in Florida has been published, while
other evaluation findings and efforts have been distributed only to a limited degree. Based on
interviews with program participants, spouses, and associates, Carter and Rudd (2000) found that
consistent program outcomes included: "networking, a broader perspective of the issues, an
increased knowledge of people's personalities, and a continued desire to learn and keep learning
throughout their life" (p. 203). A few years later, Carter and Ladewig (2004), completed an
unpublished evaluation that provided some confirmation of these findings. Program alumni were
asked to identify "outstanding features of this program that make it attractive for someone in
agriculture and natural resources." The top three findings included: (1) increased contacts and
networking opportunities, (2) exposure to different people, places, and ideas, and (3) leadership
development. In addition, study participants reported the program improved their
communication skills and caused them to be more involved in organizations (agriculture and
others) on a community, state and national level. Carter and Ladewig also asked program alumni
to identify areas for program improvement. Based on responses in questionnaires and small
group discussions, four themes were identified where the program needs to incorporate changes
and continue to monitor for progress. The four theme areas were: alumni involvement, program
design, program persona, and life after the leadership program experience. Recommendations
for changes were specific to the program, and the importance of context in the recommendations
may be why the Carter and Ladewig study was unpublished.
Grassroots Leadership Development
Scholars that study grassroots associations argue that leadership in grassroots situations
"tends to be very different from work organization management, especially business and
government organization management" (Smith, 2000, p. 150). Based on an exploration of 23
organizations involved with grassroots leaders, the W.K. Kellogg Foundation (2000) has
identified five key characteristics of grassroots leadership:
1. They have different motivations and needs than those of traditional "positional" leaders;
2. Investing in grassroots leadership development encourages long-term problem-solving;
3. In developing grassroots leaders, the best results are achieved by using a triple focus on
the individual leaders, the involved organization, and the community or issue of concern;
4. Programs work best when the investment in grassroots leadership is deliberate; and
5. Grassroots leaders encourage further support of grassroots leadership. (p. 6)
Although formal training of grassroots leaders has been limited (Smith, 2000), such efforts
have been increasing, especially since the late 1990s (Behrens & Benham, 2007). The W.K.
Kellogg Foundation (1999) has identified key results or characteristics of effective grassroots
leadership development programs. Those include:
1. Connect individuals and organizations to an overarching vision or goal for the
2. Assist leaders in moving from single issue or isolated problem-solving to an
understanding of the interrelationship of issues, power and change strategies in a
3. Expand ability to work through differing positions and across divisive boundaries to
achieve results for the community;
4. Foster networking within and outside the community; and
5. Offer follow-up to support leaders in using new skills or strategies. (W.K. Kellogg
Foundation, 1999, p. 8)
The Center for Participatory Change has proposed eight core elements for achieving
success in grassroots efforts: (1) relationships and fun, (2) mission and vision, (3) values, (4)
community work, (5) building the organization, (6) resources, (7) networks and partnerships, and
(8) growth and learning (White, Castelloe, Butterworth, & Watson, 2003). These elements
should be considered when planning grassroots leadership development programs.
Although leadership development efforts in rural and agricultural settings are in some
ways unique, they are not necessarily distinct from grassroots leadership development. Efforts to
improve leadership development programs for agricultural and rural constituents would benefit
from taking a grassroots approach and applying the findings associated with grassroots
leadership development (Burgraff, 1999; Fluharty, 2004; W.K. Kellogg Foundation, 1999). One
resource for guiding efforts to improve leadership development programming is The Handbook
for Leadership Development Evaheation, published by the Center for Creative Leadership and
edited by Hannum, Martineau, and Reinelt (2007).
Many leadership development programs are conducted through nonprofit organizations.
The nonprofit sector is thriving, and the opportunities facing nonprofit organizations present a
promising future for those that are able to adapt appropriately.
The number of Americans employed in the nonprofit sector has doubled in the last 25
years (Independent Sector, 2004, p. 1). Based on 1997-2001 figures from the United States
Bureau of Labor Statistics, the average annual growth rate for employment in nonprofits is 2.5
percent, which is well above that for business (1.8%) and government (1.6%). Total
employment in the nonprofit sector is 12.5 million, 9.5 percent of total all employment in the
United States (Independent Sector, 2004).
Based on a 2002 report published by Rollins College, Florida' s nonprofit sector is a maj or
contributor to the state economy. According to the report, Florida' s nonprofit organizations:
* Number more than 50,000;
* Directly employ approximately 430,000 people;
* Generate an additional 360,000 jobs as a result of spending by the organizations and their
* Comprise the state's sixth largest source of employment among all industry sectors;
* Hold assets exceeding $63 billion;
* Receive more than $43 billion in annual income;
* Generate more than $22 billion in total personal income;
* Generate more than $61 billion in total economic activity;
* Have grown faster than the state's overall economy; and
*Attract 88 million hours of volunteer time, equal to the work of more than 42,000 full-time
employees. (Public Sector Consultants, 2002, p. 7)
With respect to jobs in Florida, the report indicates that "approximately 6 percent of all
employment in Florida is based in the nonprofit sector" (Public Sector Consultants, 2002, p. 11).
Florida's nonprofit organizations employ more people than many other employment sectors,
including the construction; wholesale trade; transportation, communication, and utilities; and
mining sectors. Employees in Florida' s nonprofit organizations are paid an average wage of
$26,197. By applying a model developed by the U. S. Bureau of Economic Analysis, researchers
determined that 30.4 jobs are created for every $1 million in nonprofit expenditures. In total, it is
estimated that the nonprofit sector generated an additional 360,415 jobs in Florida in 1999
(Public Sector Consultants, 2002, p. 23).
One reason for the nonprofit sector' s strong growth in the United States is the success they
have had in the political arena. Although nonprofits represent only 7 percent of the interest
groups in Washington DC, from the 1960s through the 1990s nonprofit groups "accounted for
anywhere from 24 to 32 percent of the congressional testimony, generated between 29 and 40
percent of the press coverage of pending legislation, and were nearly 80 percent as effective in
passing legislation they favored as the business lobbies against which they were often arrayed"
(Salamon, 2002b, p. 43-44).
Crucial nonprofit opportunities are social and demographic shifts, new philanthropic
resources, greater visibility and policy salience, and growth of government social welfare
spending (Salamon, 2002b). Demographic shifts in the world's population have allowed
nonprofits to tap into the resources (both financial and human) of an aging population. Retirees
are now an important source from which to draw volunteers. New philanthropic resources are
another area of opportunity, including intergenerational transfer of wealth, new wealth, and
corporate partnership s/coll ab orati ons (Salamon, 2002b). In relati on to these prospects, the
financial industry has created new avenues for securing private donations, such as trusts,
annuities, etc. Nonprofit organizations must capitalize on these new philanthropic resources.
For many nonprofits, recent tragedies in the form of natural disasters and terrorist attacks have
resulted in greater visibility and policy salience. How a nonprofit organization responds to such
issues can have significant impacts on its future reputation and support. Many social welfare
programs have capitalized on opportunity areas and have received larger government
investments in recent years.
The opportunities facing the nonprofit sector present a promising future for those
organizations that are able to adapt appropriately. The nonprofit response has generally been "a
story of resilience," but it requires a balance "between the steps nonprofits must take to survive
and those they must take to remain distinctive" (Salamon, 2002a, p. 5).
One significant deficiency on the horizon for the nonprofit sector is leadership. In a
nationwide study, Tierney (2006) found that nonprofit organizations will need 640,000 new
senior-level managers in the next decade. This number is 2.4 times the number currently
employed. By the year 2016, America' s nonprofit organizations are expected to need about
80,000 new senior managers per year. According to Tierney, "the leadership deficit looms as the
greatest challenge facing nonprofits over the next ten years" (p. 4).
Although leadership in the form of paid staff is critical for most nonprofit organizations,
leadership from volunteer board members is also a topic worthy of consideration. The nonprofit
literature consistently identifies positive relationship between organizational effectiveness and
overall board performance (Carver, 1997; Drucker, 1990; Herman & Renz, 2002; Herman, Renz,
& Heimovics, 1996). Furthermore, research suggests that nonprofit organizations can take
intentional steps to improve board performance (Brudney & Murray, 1998; J. C. Green &
Griesinger, 1996; Herman & Renz, 1998, 2004; Holland, Chait, & Taylor, 1989; Holland &
Jackson, 1998). "The generally accepted view is that boards that follow the right practices or use
more of the practices will be more effective than boards that do not, and that more effective
boards will result in more effective organizations" (Herman & Renz, 1999, p. 113). This is not
to say that there is one best way to create effective boards, but it does suggest that "there are
likely to be improvements in both board and organizational effectiveness from efforts to improve
how boards do their work" (Herman & Renz, 1997, p. 14)
Holland and Jackson (1998) have identified six dimensions of board competency that they
suggest capture the foundational elements of nonprofit governance:
Contextual: the board understands and takes into account the culture, values, mission, and
norms of the organization it governs.
Educational: the board takes the necessary steps to ensure that members are well informed
about the organization, the professions working there, and the board's own roles,
responsibilities, and performance.
Interpersonal: the board nurtures the development of its members as a group, attends to the
board's collective welfare, and fosters a sense of cohesiveness and teamwork.
Analytical: the board recognizes complexities and subtleties in the issues it faces, and it
draws upon multiple perspectives to dissect complex problems and to synthesize
Political: the board accepts that one of its primary responsibilities is to develop and
maintain healthy two-way communications and positive relationships with key
Strategic: the board helps envision and shape institutional direction and helps ensure a
strategic approach to the organization's future. (p. 122-123)
These dimensions of board competence are a helpful starting point for understanding
nonprofit board performance in general, but they should be considered in context. Most of the
research on nonprofit board development and organizational effectiveness highlights charitable
organizations with paid staff members, leaving out membership organizations and volunteer
managed organizations. Smith and Shen (1996) conducted a study of thirty-nine volunteer-
managed organizations and found that "many hypotheses suggested by others for nonprofit
organizations with paid staff do not seem to transfer to volunteer nonprofit groups" (p. 271).
Smith and Shen did find that volunteer nonprofit organizational governance significantly effects
reputational effectiveness. However, they urge additional research with volunteer nonprofit
organizations, and they encourage the investigation of other measures of effectiveness.
More empirical evidence is needed to determine if the overriding hypotheses on nonprofit
organizational effectiveness are applicable in nonprofit membership organizations.
Unfortunately, "the more general tendency of nonprofit scholars is to ignore membership
organizations, especially grass-roots associations" (Smith & Shen, 1996, p. 285). The Florida
Farm Bureau Federation (FFBF) is such an organization and provides the context for this
Florida Farm Bureau Federation
The Florida Farm Bureau Federation (FFBF) is a nonprofit organization that is closely
associated with grassroots, agricultural, and rural leadership development. A program conducted
by FFBF for its local leaders is the focus of this study.
In the 1800s, the earliest organizations of farmers began to develop, organizing under a
variety of names and philosophies, including The Grange, The Farmer'sFFF~~~~FFF~~~FFF Alliance, The
Agricultural Wheel, The Ancient Order ofGleaners, and the Equity. The Hatch Act of 1862
established the land grant university system and agricultural experiment stations. The
Cooperative Extension Service soon provided agricultural agents in each county (American Farm
Bureau Federation, 2005).
Continued challenges in the agricultural sector encouraged the formation of county farm
bureaus in the early 1900s. The first representation was in 1911 when John Barron was hired as
a "farm bureau" representative within New York' s Broome County Cooperative Extension
Service. Missouri formed the first statewide Farm Bureau organization in 1915, and the
American Farm Bureau Federation (AFBF) was organized in 1919 (AFBF, 2005).
In 1941, Florida Citrus Growers (FCG) contacted AFBF in Chicago and AFBF sent a
representative to the next FCG directors meeting. A general farmers meeting was called in
Orlando, attended by "Cap'n Ed" O'Neal, president of AFBF, and it was decided that FCG would
be laid to rest and resurrected as the heart of a new Farm Bureau organization affiliated with
AFBF. The first FFBF convention was held on November 15, 1941. About 100 farmers
attended and elected George L. Fullerton as FFBF's first president. The charter outlined the
goals of the Federation: (1) give Florida farmers a special identity; (2) study and promote better
cultural and research practices, product quality improvement, improved marketing methods and
market stabilization; (3) represent farmers in the Legislature, (4) compile and disseminate
agricultural and market data to farmers; and (5) organize county Farm Bureau units. On March
13, 1942 the first county Farm Bureau was formed in Dade County. By the time Florida Farm
Bureau celebrated its first birthday, membership had reached 1,180 and 17 county Farm Bureaus
had been established. By the end of the 1950s, membership in FFBF had more than tripled.
More than half of FFBF counties had their own offices, many staffed with full-time secretaries
Mission, Vision, and Goals
The FFBF mission is "to increase the net income of farmers and ranchers, and to improve
the quality of rural life" (personal communication, Rod Hemphill, November 9, 2005). The
FFBF board of directors have adopted the following vision: "Florida Farm Bureau will be the
most effective, influential and respected Farm Bureau in the nation and will be recognized as
Florida's 'Voice of Agriculture'" (FFBF, 2006a). The mission of the American Farm Bureau
Federation is "to implement policies that are developed by members and provide programs that
will improve the financial well-being and quality of life for farmers and ranchers" (American
Farm Bureau Federation, 2006).
The FFBF Goals, as outlined by the elected leadership, include:
1. Continue to increase membership and maintain a strong financial base;
2. Continue to build stronger relationships with the county Farm Bureaus;
3. Have a strong PR program to promote Farm Bureau and the value of agriculture to the
citizens of Florida;
4. Have our members, policy makers, agricultural organizations and governmental agencies
view Farm Bureau as a dependable source of factual, reliable and timely information on
5. Continue to encourage Farm Bureau members to become involved in the political process
and seek public office;
6. Strengthen our committee system Women, Young Farmer & Rancher, Advisory
Committees, Ag in the Classroom, etc. and increase opportunities to involve more
people in Farm Bureau;
7. Work closely with the other agricultural groups, IFAS and the regulatory agencies;
8. Keep up our lobbying efforts in Tallahassee and increase our efforts in Washington;
9. Provide educational programs for members, county leaders, staff & agents; and
10. Seek ways to provide added services and benefits to our members. (Loop, 2004)
Local Farm Bureau Units
According to the FFBF (2006b) County President' s Desk Book, "the county Farm Bureau
is the most important unit of Farm Bureau" (p. 68). These local groups are organized to:
1. Solve the problems of the farm, the farm home, and the rural community through the
advantages of organized action. Why? So that those engaged in agriculture may have the
opportunity for economic prosperity in their chosen work.
2. Cultivate free enterprise and citizenship by instilling in Americans those ideals and
principles which are the foundations of our agricultural heritage.
3. Help the Florida Farm Bureau Federation, the Florida Department of Agriculture and
Consumer Services, and the Agricultural Extension Service develop better methods and
practices in farming and farm management, thus improving the conditions of rural life.
4. Unite the farmers of the county into an organization for the promotion and protection of
their common interests without regard to political or religious affiliation.
5. Foster and encourage the development of commodity marketing, and the purchase of
farm supplies on a cooperative non-profit basis.
6. Cooperate in the achievement of common aims and purposes with other county Farm
Bureaus, the Florida Farm Bureau Federation, and the American Farm Bureau
Federation. (FFBF, 2006b, p. 68)
These local tasks are intended to be carried out through two formal structures: the Farm
Bureau committee system and the Board of Directors (FFBF, 2006b). Although county Farm
Bureaus may involve these structures separately, the local committee system is often embedded
within the Board of Directors. As a result, efforts for change need to be channeled through the
directors on the local board.
Local Boards of Directors
"The success of a county Farm Bureau depends largely upon the interest shown by its
Board of Directors" (FFBF, 2006b, p. 51). In this capacity, FFBF has over 700 local board
members, grouped into 61 local boards (FFBF, 2006c). These board members are elected by the
local membership and assigned the task of providing leadership for local Farm Bureau initiatives
and issues. Although the local boards vary considerably in size and structure, some
generalizations about board members can be drawn from Carter's 2004 study of FFBF. "On
average, Farm Bureau board members spend 8.4 hours per month on Farm Bureau activities,
which includes participating in meetings, activities, events and conventions and reading
information in support of these activities" (Carter, p. 86). FFBF local board members attend an
average of one Farm Bureau event per month. In 2004, the average tenure of FFBF local board
members was 1 1.5 years, with 14% having served more than 20 years. One-third of the board
members had served as president of the local FFBF board. They were also active in other
organizations, with 71% actively involved in other agricultural organizations and 64% actively
involved in other civic organizations. More than half of the FFBF board members were serving
on boards of other organizations. The average length of Farm Bureau membership for local
FFBF board members was more than 20 years. Nearly three-fourths of these board members
were raised by parents who were also Farm Bureau members. The maj ority were male (88%),
married (86%), over 45 years old (54%), and had children (94%).
Board member selection and rotation varies from one local board to another. Some local
boards have specific guidelines and restrictions while other board members
FFBF field staff members are the paid employees of the state federation who serve as
liaisons between the county Farm Bureau unit and the state organization (Carter, 2004; FFBF,
2006a). FFBF divides Florida into eight geographic districts, and one field staff member is
assigned to each district. FFBF field staff members serve in the official capacity of "Assistant
Director of Field Services." Accordingly, field staff members are "responsible for working
closely with the County Farm Bureau unit to accomplish goals and obj ectives of the Florida
Farm Bureau Federation" (FFBF, 2005a, p. 1). They generally meet on a monthly basis with
each of the local FFBF boards within their district. Field staff members provide support in the
implementation of Farm Bureau programs and activities.
Prior to the Strengthening the Voice" program the FFBF field staff members' role in
leadership development lacked any formal structure. However, a traditional responsibility of the
field staff has always been to surface grassroots leaders for the organization. The actual methods
of recruiting and training these leaders varied from one field staff member to another (personal
communication, Ray Crawford, February 2, 2007).
Delivery of the Farm Bureau Found'ations program was shared by field staff and other
FFBF staff. The preparation for this delivery included a three-day train-the-trainer session. In
addition to this preparation, FFBF field staff members are in the best position to provide
obj ective observation of changes in local board behavior, including implementation of the Farm
Bureau Found'ations Best Practices. The primary advantage of the field staff members is their
regular interaction with local FFBF boards and members.
Chapter 2 provided the theoretical and conceptual framework for the study. Goal setting
theory was highlighted, and it was connected with the belief that conscious goals affect action.
Goal specificity and difficulty are important goal components that affect outcomes. According
to goal setting theory, goals affect performance through four mechanisms: (1) choice/direction,
(2) effort, (3) persistence, and (4) strategies. In addition, goal performance is explained by three
moderators: (1) goal commitment, (2) feedback, and (3) task complexity. Findings on the effect
of demographic variables on goal outcomes have not been consistent. However, some findings
suggest the need to control for education level, ethnicity, job tenure, age, and gender.
Adult education literature was discussed, including research on adult motivation to learn.
W10dkowski (1999) suggests that adult motivation to learn is the sum of four factors: success,
volition, value, and enjoyment. Appropriate program planning models for adult education
include: the Andragogy in Practice Model (Knowles et al., 2005) and the Conceptual
Programming Model (Boone et al., 2002). Kirkpatrick' s (1998) Four Levels is helpful for
evaluation, but opportunities to calculate return on investment should also be considered.
Research Eindings related to rural leadership development were discussed. Published
research specific to rural leadership development is limited. However, an evaluation of the
Wisconsin Rural Leaders program suggests that "rural leaders learn best by a process of action
and reflection" (Dhanakumar et al., 1996, p. 6). A model for rural leadership development
programming advocated for activities that stimulate critical thinking. Research on agricultural
leadership development programs hold relevance for rural leadership development, but few
published studies exists. Studies from the Kellogg Foundation and the Center for Participatory
Change provide guidance for grassroots leadership development programs, and these Eindings
have application in the area of rural leadership development.
Many leadership development programs are conducted through nonprofit organizations.
The nonprofit sector is thriving, and the opportunities facing nonprofit organizations present a
promising future for those that are able to adapt appropriately. A Florida nonprofit organization
that is closely associated with grassroots, agricultural, and rural leadership development is the
Florida Farm Bureau Federation (FFBF). FFBF is the focus organization of this research study.
Delivery of the Farm Bureau Foundations training program was shared by FFBF field staff and
other FFBF staff. Recipients of the training were FFBF local board members. Job duties and
demographics of these groups were discussed.
Figure 2-1. Theory of planned behavior. Note. From "TpB Model," by I. Ajzen, 2006, Online
at http ://people.umass.edu/aizen/tpb .html. Copyright 2006 by Icek Aizen. Reprinted
Goals and Purposes for Learning
Z ~individual and Situational Differences
i ~Core Adult Learning Principles \
1. Learners Need to Know
2. Self-Concept of the Learner
3.Prior Experience of the Learner a)
4. Readiness to Learn T
9 life related ~
R 7 5. Orientation to Learning
6. Motivation to Learn
Individual Learner Differences
Figure 2-2. Andragogy in Practice Model. Note. From "The Adult Learner," by M. S. Knowles,
E. F. Holton, and R. A. Swanson, 2005, p. 149. Copyright 2005 by Elsevier.
Reprinted with permission.
Although the study involved collection of qualitative support data, the overall design of the
study was primarily quantitative. Quantitative research starts with a more specific problem than
qualitative research; it is generally considered deductive in nature, used to test a preexisting
theory. Quantitative research is characterized by its use of obj ective measurement and statistical
analysis of numeric data (Ary et al., 2002).
Quantitative research involves the use of standardized instruments. Although preexisting
instruments can be quite useful, especially in providing comparison data, often social science
researchers develop quantitative instruments that are designed to measure more specifically the
variables involved in a particular situation. At the time of this study, no standardized
instruments had been developed for testing goal-setting theory in a universal setting (personal
communication, Edwin Locke, May 4, 2006). The researcher and proj ect administrators
developed three different questionnaires to collect data for this study.
Ary et al. (2002) describe a survey as a research technique in which data are gathered by
asking questions of a group of individuals. Survey research asks questions about the nature,
incidence, or distribution of variables and/or the relationships among those variables. No
manipulation is attempted on the variables, only descriptions of variables and their relationships
as they naturally occur.
Descriptive research was used to accomplish research obj ectives one through three:
1. Describe select demographics of local Florida Farm Bureau board members and program
2. Measure the engagement level of local Florida Farm Bureau board members in FarmFFFFF~~~~~~~FFFFFF Bureau
Foundations Best Practices; and
3. Measure the goal attributes experienced by program participants with their Farm Bureau
Foundations Best Practices goal.
According to Ary and others (2002), descriptive research is used to "summarize, organize, and
describe observations" (p. 118). The descriptive statistics relevant to objectives one through
three will be summarized by arranging the measures into frequency distributions and presenting
them in graphic form.
A correlational and causal-comparative or ex post facto design was employed to
accomplish obj ectives four and five:
4. Explain individual-level performance with best practices identified in the FarmFFFFF~~~~~~~FFFFFF Bureau
Foundations program; and
5. Explain board-level performance with best practices identified in the Farm Bureau
In an ex post facto design, the researcher does not have direct control over the independent
variable(s) and any variations in the independent variable(s) have already been determined prior
to the research being conducted (Ary et al., 2002).
With respect to obj ectives four and five, the dependent variable was Farm Bureau
Foundations Best Practices performance. Performance at the individual level was self-reported,
whereas performance at the board-level was evaluated as the mean of ratings among local board
members. In addition, board performance was rated by a FFBF staff member familiar with the
local board. A more detailed description about how performance was measured and indexed is
provided later in this chapter.
Independent variables analyzed for obj ectives four and five were demographic and
program-related variables. The demographic variables included: county of residence, age,
gender, ethnicity, education level, and Farm Bureau tenure. Program-related variables included:
program participation, perceptions of the goal-setting process, and perceptions of engagement in
This study was conducted over the course of one year. Figure 3-1 outlines a complete
timeline for the procedures involved in this study.
This study was a census of two separate groups: FFBF local board members and FFBF
field staff. Although the local board members were a complete population in and of themselves,
they were further divided into sub-groups based on participation in the Farm Bureau
Foundations program. During this study, different survey instruments were administered to each
population and/or sub-group. Data from this study were analyzed at two different levels of unit
analysis: individual and local board of directors. The following sections describe the selection of
individuals for participation in the study. The response rates are described in the data collection
procedures section of this chapter.
Participant Selection: FFBF Local Board Members
FFBF has over 700 local board members serving 61 county-level membership groups
(FFBF, 2006c). These local board members were the target population for the Suengthenll~lingl the
Voice program. This research study includes all FFBF local board members because program
participants and non-participants alike can provide valuable information on best practices
adoption. In order to compare effects of program participation, the population of FFBF local
board members was divided into three sub-groups. The first sub-group, labeled the "treatment
group," included local FFBF board members who participated in the Farm Bureau Found'ations
program. The second sub-group, labeled the "control group," included board members in
counties where no one participated in the FarmFFFFF~~~~~~~FFFFFF Bureau Found'ations program. The third sub-
group, labeled the "spillover treatment group," included board members who did not participate
in the Farm Bureau Found'ations program but were members of a local FFBF board in which
another director participated in the program. (Note: Local boards are generally identified by
county, with the exception that Palm Beach County has two local boards Palm Beach and West
For the treatment group, a list of FarmFFFF~~~~~~~FFFFFF Bureau Found'ations program participants was
obtained from the FFBF Field Services office. The participant list included contact information
and was used as the population frame. Some of the individuals on the list could not be matched
to FFBF's current list of local board members. However, the researcher made the decision to
include everyone on the participant list rather than exclude individuals who fit the true
population yet were missing from the FFBF list of current local board members. The initial list
of participants identified 121 program participants. However, additional investigation and
communication with program presenters suggested that program participation totaled 156. Of
those 156, though, not all were local FFBF board members. FFBF staff members were able to
provide contact information for 138 program participants. An invitation to participate in the
study was mailed to those individuals on the updated program participation list (n=138).
Initially, the control group and the spillover treatment group were combined based on their
common attribute their absence from participation in the FarmFFFFF~~~~~~~FFFFFF Bureau Found'ations workshop.
A database of local board members was obtained from FFBF and identified workshop
participants were removed from the list. The remaining 611 local board members represented
the most comprehensive list available to serve as the population frame for local FFBF board
members who did not participate in the FarmFFFFF~~~~~~~FFFFFF Bureau Found'ations workshop. This aggregate of
the control group and the spillover treatment group is referred to as the non-attendee group
(n= 61 1).
Although the initial division of the population by workshop attendance was considered
accurate, based on Farm Bureau' s records, survey evidence collected during the course of this
study revealed some discrepancies. From the non-attendee group, 10 survey respondents
revealed that they had in fact participated in the workshop. In addition, one survey respondent
from the treatment group indicated absence from the workshop. Other information collected
through letters and phone calls revealed that some individuals were duplicated or were not in the
target population. Specifically, one duplicate was removed from the treatment group and three
duplicates were removed from the non-attendee group. In addition, three individuals from the
non-attendee group were identified as deceased. From the treatment group, one individual was
dropped after contacts questioned the accuracy of the name on record, and another was dropped
after no credible contact information was obtained. Responses from 22 individuals revealed that
they either were no longer members of the local FFBF board or served in an ex-officio capacity
and therefore did not meet the criteria for study participation. Among those non-board-members,
six were from the treatment group and sixteen were from the non-attendee group.
After the larger population of FFBF local board members was divided by workshop
attendance, the distinction between the control group and the spillover treatment group was made
using the board member database provided by FFBF. Once the database was sorted by county,
board members in counties that had a participant in the Farm Bureau Found'ations program were
removed from the list. The remaining individuals in FFBF's list of board members represented
the most comprehensive list available to serve as the population frame for the control group. For
the spillover treatment group, board members in counties that had participants in the program
were identified, and then actual participants were removed from that list. The resulting
population frame was the most comprehensive list available to represent the population of the
spillover treatment group. A final summary of the sub-groups is presented in Table 3-1.
Participant Selection: FFBF Field Staff
Because of their regular interaction with local FFBF boards, the FFBF assistant directors
of field services (a.k.a. field staff) are in the best position to observe changes in local board
behavior, including implementation of the FarmFFFFF~~~~~~~FFFFFF Bureau Found'ations Best Practices (personal
communication, Ray Crawford, February 12, 2007). FFBF divides Florida into eight geographic
districts, and one field staff member is assigned to each district. On average, field staff members
work with eight different counties. The researcher sought a census of FFBF field staff members
in order to obtain a perspective of local FFBF board behavior relative to other local FFBF
boards. The FFBF Field Services office supplied contact information for the FFBF field staff
members, and all were invited to participate in the study.
The study used three different self-administered mail questionnaires to collect information
relevant to the study objectives. The researcher and project advisory committee worked
collaboratively to design all three instruments. The "Local Board Practices Survey" (Appendix
D) assesses local FFBF boards of directors' current and recent application of FarmFFFF~~~~~~~FFFFFF Bureau
Foundations Best Practices. The "Best Practices Goal Survey" (Appendix E) assesses the FarmFFFFF~~~~~~~FFFFFF
Bureau Found'ations Best Practices adoption and goal achievement by Farm Bureau
Foundations program participants. The "Field Staff Perceptions Survey" (Appendix F) collects
information from FFBF field staff regarding their observations of Farm Bureau Found'ations
Best Practices being implemented by local FFBF boards of directors.
The reliability and validity of the instruments are important aspects to consider in all
research studies utilizing survey instruments. Validity is a concept that refers to whether or not
an instrument measures what it is intended to measure (Ary et al., 2002). In an effort to improve
validity of the researcher developed instruments, a panel of experts reviewed each instrument.
This panel included individuals with research expertise at the University of Florida and
individuals with organizational expertise in FFBF. The researcher applied feedback from the
panel of experts and made appropriate revisions to the survey instruments.
The researcher pilot tested the survey instruments using cognitive interviews. Cognitive
interviewing is not without its limitations. However, there is general agreement in the social
sciences that cognitive interviewing is a valuable tool for testing and evaluating survey questions
(Presser et al., 2004). For this study, five FFBF state board members served as the subjects in
the cognitive interviewing process. The protocol applied in the cognitive interviews can be
found in Appendix G. The researcher used feedback from the cognitive interviews to further
strengthen the validity and reliability of the instruments. The most significant concern that
surfaced during the cognitive interviews related to survey length. As a result, the researcher
eliminated a few questions that were deemed unnecessary. In addition, a change was made in the
numbering of the questions. Rather than using consecutive numbers throughout, questions were
identified by sections and numbering was restarted at the beginning of each section. Although
the change in numbering did not actually reduce the practical length of the survey, it may have
provided a reduction in the perceived length. Additional research is needed to validate the use of
As an additional effort to strengthen and provide support for the validity of the survey
instruments, the researcher asked his PhD committee to review the Einal instruments before
mailing them to the study population. During this step, no additional changes to the instruments
Local Board Practices Survey
The "Local Board Practices Survey" (Appendix D) is a researcher-developed instrument
designed for use as a mailed questionnaire to local-level FFBF board members. The instrument
evaluates local FFBF boards of directors' current and recent application of FarmFFFF~~~~~~~FFFFFF Bureau
Foundations Best Practices. The "Local Board Practices Survey" has four sections:
A) "Board practices during the past six months...,"
B) "Comparison with other six-month periods...,"
C) "Limits to engagement in board member during the past six months...," and
D) "Please tell us about yourself..."
Section A includes both individual and board-level assessment of 17 local Farm Bureau board
practices. These practices were taken directly from the Farm Bureau Found'ations curriculum.
Section B includes two questions that ask participants to identify to what degree their individual
engagement-level and the local board's individual engagement-level has been more or less than
usual. Section C includes four questions related to confounding factors that may have limited
individual engagement in board practices. Specifically, the questions ask respondents to identify
limits to engagement due to personal and professional factors. Section D includes nine
demographic-related questions, including county of residence, age, gender, ethnicity, education-
level, length of membership in FFBF, FFBF board tenure, recent meeting attendance, and Farm
Bureau Found'ations workshop attendance. The question about workshop attendance is included
to check for discrepancies in population sub-grouping. In addition to the specific questions,
space is provided in the questionnaire for study participants to add additional comments and
Best Practices Goal Survey
The "Best Practices Goal Survey" (Appendix E) is a researcher-developed instrument
designed for use as a mailed questionnaire to Farm Bureau Found'ations program participants
(the treatment group). The instrument is intended to be used to evaluate the relationship between
goal-setting theory components and application of the Farm Bureau Found'ations Best Practices.
The "Best Practices Goal Survey" includes the same sections as the "Local Board Practices
Survey," and it adds an additional section: "Best practices goal setting process..." This
additional section includes 24 questions related to constructs associated with goal setting theory
and the high performance cycle. Constructs represented include: goal specifieity, goal difficulty,
goal commitment, goal-related self-efficacy, and performance. This section is appropriate for
Farm Bureau Found'ations program participants because they set individual goals for best
practice adoption and should be able to reflect on that process. This instrument is not
appropriate for those who did not participate in the program since they are not likely to have a
goal on which to base their answers. In addition, the demographic section of the survey
instrument asks participants to indicate the current availability and use of materials distributed at
the FarmFFFFF~~~~~~~FFFFFF Bureau Found'ations workshop.
Field Staff Perceptions Survey
The "Field Staff Perceptions Survey" (Appendix F) is a researcher-developed instrument
designed for use as a mailed questionnaire to FFBF Hield staff. The instrument is intended to be
used as an external assessment of local FFBF boards of directors' current and recent application
of Farm Bureau Found'ations Best Practices. In addition, the instrument collects data on the
Hield staff member' s tenure with Farm Bureau and their overall perceptions of local board
members' general application of Farm Bureau knowledge. The researcher designed the "Field
Staff Perceptions Survey" for assessment of local boards where directors completed the FarmFFFFF~~~~~~~FFFFFF
Bureau Found'ations program and also of local boards where no directors participated in the
FarmFFFFF~~~~~~~FFFFFF Bureau Found'ations program. The external assessment of participants and non-
participants allows for additional comparison data for evaluating the impact of the Farm Bureau
Foundations program. The "Field Staff Perceptions Survey" includes 17 questions related to
local board application of FarmFFFFF~~~~~~~FFFFFF Bureau Foundations Best Practices. These questions are
repeated for each local board the Hield staff member oversees. Space is provided for study
participants to provide additional comments and qualitative information.
Data Collection Procedures
Institutional Review Board Approval
Prior to data collection, the Institutional Review Board (IRB-02) at the University of
Florida reviewed the study procedures and evaluated the ethical soundness of the study. Copies
of the informed consent letters for study participants and a proposal to conduct the study were
submitted to the IRB-02. The informed consent letters described the study and the voluntary
nature of participation (Appendix I). The IRB-02 approved the research protocol and assigned
an IRB protocol number: 2006-U-824. Once approval was granted, data were collected and
analyzed by the researcher.
Collection Procedures with FFBF Local Board Members
The contact procedures for this study employed strategies for achieving high response
rates. Czaj a and Blair (2005) have identified a number of researcher practice aspects that
improve unit non-response: prior notification about the survey, efforts to reach the respondent,
initial contact and respondent selection intermediary (gatekeeper), requesting participation,
follow-up efforts, and refusal conversation (p. 199). These practices are reflected in Dillman's
(2000) Tailored Design M~ethod. Specific elements of the method include: "(1) a respondent-
friendly questionnaire, (2) up to five contacts with questionnaire recipient, (3) inclusion of
stamped return envelopes, (4) personalized correspondence, and (5) a token financial incentive
that is sent with the survey request" (p. 150). Dillman recommends financial incentives of one to
five dollars be included with requests to respond to a mailed questionnaire. In the case of Farm
Bureau, such an incentive could be perceived by study participants as a waste of their Farm
Bureau membership dollars. As an alternative, FFBF planned to include a lapel pin in the
mailings. Unfortunately, the design and production of the pins took longer than FFBF
anticipated, so the pins were not available for inclusion in the survey mailings. Consequently,
Dillman's fifth element recommended for improving response rates was not implemented.
The researcher collected data during October and November of 2006. In an effort to
improve credibility of the study, FFBF coordinated all mailings to participants, using FFBF
stationary and envelopes. On October 6, 2006, FFBF mailed a brief a pre-notice letter to study
participants. The letter was sent from the FFBF president and indicated to recipients that they
would soon be receiving survey materials. As suggested by Dillman (2000), the letter was
"brief, personalized, positively worded, and aimed at building anticipation rather than providing
the details or conditions for participation in the survey" (p. 156). The template for the pre-notice
is in Appendix H.
Three days after the pre-notice letter was mailed, the questionnaire packets were mailed to
participants. The cover letter template for the survey mail-out was the informed consent letter
approved by IRB-02. The cover letter included critical elements identified by Dillman (2000) to
motivate response behavior:
* Inside name and address,
* What is this letter about?,
* Why this request is useful and important,
* Answers are confidential,
* Voluntary participation,
* Enclosure of stamped return envelope, and
* Who to contact with questions.
Due to the volume of mailings, real signatures were not applied. However, Dillman (2000)
offers that personalization, including use of individual names and letterhead, "may compensate
to some extent for being unable to apply real signatures" (p. 165). The template applied for the
questionnaire cover letters is displayed in Appendix I.
Seven days after the questionnaire mail out, post cards (Appendix J) were mailed to
participants, thanking them for their participation and reminding them to return their completed
questionnaires. Such a postcard is recommended by Dillman's (2000) TailoredDesign Method.
Research by Dillman, Clark, and Sinclair (1995) suggests that use of reminder postcard increases
total response rates by eight percentage points.
At the time of the initial survey mailing, each study participant was assigned an individual
identification number and all instruments were coded with these individual identification
numbers. As the researcher received the completed instruments, the identification numbers were
used to eliminate respondents from future requests for participation.
Two weeks after the reminder postcards were mailed, replacement questionnaires were
mailed to study participants from whom a completed questionnaire had not yet been received.
As encouraged by Dillman' s (2000) TailoredDesign method, the cover letter for this fourth
contact expressed increased urgency and brought a tone of insistence beyond that of the previous
contacts. In addition to components included in the cover letter of the initial questionnaire mail
out, the cover letter for the replacement questionnaire included feedback indicating others have
responded, yet we haven 't heard fiom you. These feedback additions are consistent with
Dillman' s TailoredDesign method. The template applied for the replacement questionnaire
cover letters is displayed in Appendix K.
Dillman (2000) recommends that researchers follow up with survey non-respondents for a
fifth contact, invoking special procedures. The insistent nature and effectiveness of this request
"stems from the simple fact that it is a fifth request, and it is being made by special mail or by
telephone" (Dillman, p. 184). Certified mail, special delivery, and telephone follow-up phone
calls are all effective ways to handle this "invoking of special procedures." Research by Moore
and Dillman (1980, cited in Dillman, 2000, p. 186) suggests that any one of these methods has
return advantages over sending another first-class mailing. However, these special procedures
also come with an increased cost, and that cost noticeably increases for every non-respondent
included in this fifth contact.
For this study, the fifth contact was a telephone follow-up with randomly selected non-
respondents. Rather than urge these non-respondents to find and complete the mailed survey, the
researcher asked them to answer a subset of questions over the phone. The primary use for this
telephone collected data was to investigate the potential for unit non-response bias with the
mailed questionnaires. Miller and Smith (1983) refer to this technique as a "double-dip" of non-
respondents (p. 48). The script applied in the telephone follow-ups for this study is found in
For the phone follow-ups, the researcher used SPSS@ 12.0 for WindowsTM to randomly
select FFBF Local Board Members who had not responded within four weeks after the
replacement questionnaires were mailed. Non-respondents who had participated in the Farm
Bureau Found'ations program component were over-sampled so that half of the telephone
contacts would be with program participants and half would be with program non-participants.
This over-sampling of program participants was done because the program participants were the
group of primary interest in this study. In order to obtain phone numbers for the selected non-
respondents, the researcher referenced the FFBF membership database and used the "reverse
address phone number look-up" tool on WhitePages.com. Initially, 60 non-respondents were
randomly selected for phone follow-ups. However, a high rate of failed call attempts prompted
the researcher to randomly select an additional 30 non-respondents. After two weeks of phone
call attempts, with multiple attempts for each individual, the researcher had telephone-collected
data from 32 FFBF Board Members, 15 workshop attendees and 17 non-attendees.
In addition to the value of questionnaire responses, the phone calls in this fifth contact
helped to highlight reasons for non-response to the mailed questionnaires. In some cases, an
individual shared over the telephone that they do not respond to any surveys. (Surprisingly, one
person who indicated that she never participates in surveys was willing to answer questions over
the phone.) In other cases, telephone calls revealed that the person identified for participation in
the study is outside of the target population because he/she is no longer a member of the local
FFBF board. Some telephone respondents shared that they thought they had already returned the
questionnaire. After further investigation, the researcher determined that some respondents had
returned a questionnaire for a different survey being conducted by FFBF. Unfortunately, this
other FFBF survey overlapped the survey discussed in this dissertation. It may be that a
significant portion of non-respondents confused the two surveys, which in turn impacted the
response rate for this study.
Re-group and additional mailing
As data were collected, responses suggested that 41 study participants had participated in
the FarmFFFFF~~~~~~~FFFFFF Bureau Foundations program component, even though their names were not on
FFBF's list of participants. As a result of pre-mailing grouping, these program participants were
completing the shorter questionnaire, the "Local Board Practices Survey," rather than the "Best
Practices Goal Survey" that included additional questions for program participants. In order to
obtain these additional data, the researcher conducted a follow-up mailing with these survey
respondents. The cover letter asked participants to complete the incomplete questions of the
enclosed "Best Practices Goal Survey," which was pre-filled with their originally submitted
answers to the "Local Board Practices" survey. This follow-up with 41 potentially mis-grouped
study participants yielded responses from 16 individuals. Ten of those individuals provided at
least a portion of the additionally requested data. The other six responding individuals indicated
that they had made an error on their previous questionnaire and were not in fact in attendance at
a FarmFFFFF~~~~~~~FFFFFF Bureau Foundations workshop. This evidence, in combination with FFBF's
participation records, suggested that regrouping was appropriate for only those who provided
questionnaire data in response to the additional mailing of the "Best Practices Goal Survey."
Individuals who reported program participation on the "Local Board Practices Survey" but did
not return the "Best Practices Goal Survey" were kept in their original group of non-attendees.
Unit response rate
In the end, the data collection procedures in this study yielded 85 responses from the
treatment group (n=138). This corresponds to a unit response rate of 61.6% for the treatment
group. The researcher obtained 82 responses from the control group (n=138), for a response rate
of 59.4%. The researcher obtained 244 responses from the spillover treatment group (n=442),
for a response rate of 55.2%. For the larger population of all FFBF local board members
(N=718), survey responses were obtained from 411 individuals for an overall response rate of
57.2%. When considering the local board as a level of unit analysis, responses were received
from at least one board member in 60 of the 61 local boards. This corresponds to a local board
response rate of 98.4%. The percent of counties reaching increasing levels of individual
response is displayed in Figure 3-2.
Additional data from FFBF
The mailed questionnaires asked respondents to identify their county and the number of
years they had been a Farm Bureau member. When respondents did not answer these questions,
the researcher worked with FFBF staff to obtain this information from FFBF membership
records. In order to save time, these questions were not asked in the telephone calls to non-
respondents. Instead, the researcher relied directly upon the FFBF membership records to obtain
data on respondents' county and years of Farm Bureau membership.
Although the membership records were a reliable source of information, some concerns
exist regarding the validity of the years of membership data. For less tenured members, the
FFBF database identified a specific date joined. However, for members who originally joined
before the current computer system was in place, no date joined was available. Instead, FFBF
staff estimated the years of membership based on years of paid dues, as recorded in FFBF
records. This would be a valid estimate of years of Farm Bureau membership for many study
participants. However, FFBF staff shared that if a member was late in paying their annual dues,
the number of years paid would start back at zero, causing an invalid estimate of years of
membership. This error was noticeable for respondents who reported their number of years on
the local FFBF board to be greater than FFBF's record of consecutive years of dues paid. In
these observations, the researcher increased the years of membership to the equivalent number of
years served on the local FFBF board. While this recoding was likely an under-estimation of
actual length of membership, it was the best estimation available to the researcher.
FFBF records were also used to provide information about the size of each local board.
Board size was determined manually by counting the number of board members identified with
each local board. This information was then available as an independent variable in the data
analy si s.
FFBF Field Staff
On November 3, 2006, the researcher E-mailed the FFBF assistant directors of field
services to request a brief phone interview. The template used for the E-mails may be found in
Appendix M. Attached to the E-mail was a copy of the appropriate IRB informed consent form
(Appendix N). As field staff members responded to the E-mail, the researcher scheduled and
conducted brief telephone interviews. The intent of these interviews was twofold: (1) to collect
rich qualitative data regarding field members' perceptions of the program and its impacts, and
(2) prompt field staff members to begin thinking about program related board practices, so that
they would be prepared to supply valid responses to the "Field Staff Perceptions Survey." Each
interview included two broad questions:
1. In what ways has board behavior changed or been reinforced as a result of participation in the
Socligthenl~lila the Voice program?
2. In what ways do you believe the Sor cigthenl~lila the Voice could be improved?
At the conclusion of each interview, the researcher followed up with an E-mail that
included the "Field Staff Perceptions Survey" as an attachment. The E-mail asked the field staff
member to complete each section of the questionnaire and return it to the researcher by the end
of the month. One field staff member promptly completed the questionnaire and returned it by
mail. During a FFBF field staff meeting at the end of the month, other field members said that
they never received the questionnaire. Hard copies of the questionnaire were immediately made,
and these field staff members completed the questionnaires at that time.
In November 2006, when data were being collected from FFBF field staff members, one
field staff person was on extended medical leave and another field staff position was vacant. For
both of these circumstances, the FFBF director of field services indicated that no one on staff
was qualified to answer questions about county-level board practices in the individual district
(personal communication, Ray Crawford, November 20, 2006). The researcher did obtain
survey responses from FFBF Hield staff members representing the six other districts. Data from
six of the eight regions represents a response rate of 75%. When considering the local board as a
level of unit analysis, data from FFBF Hield staff members provided ratings for 43 of the 61 local
boards. This corresponds to a response rate of 70.5% at the local board unit-level of analysis.
Data Analysis Procedures
Data collected from the survey instruments was analyzed using SAS@ 8.2 and 9. 1.2 for
According to Dillman (2000), "non-response error occurs when a significant number of
people in the survey sample do not respond to the questionnaire and have different
characteristics from those who do respond, when these characteristics are important to the
study"(p. 10). Anytime a response rate falls below 100%, a non-response bias may exist and can
be a threat to the external validity of a study (Lindner, Murphy, & Briers, 2001). When evidence
of non-response bias exists, caution must be exercised in generalizing findings beyond those who
fully participated in the study.
In order to test for unit non-response bias in data from the mailed questionnaires, the
researcher conducted tests of ANOVA to identify differences in mean responses between the
data collected from the mailed questionnaires and data collected from the telephone follow-up
interviews. Specific demographic variables considered in this comparison included: age, gender,
level of education, year of membership in FFBF, years of service on the local FFBF board, and
number of local FFBF board meetings attended in the last six months. Tests for ANOVA
showed no significant different between the two groups on the variables of age, gender, level of
education, years of service on the local FFBF board, and number of local FFBF board meetings
attended in the last six months. However, ANOVA revealed that significant differences existed
between mail respondents and phone respondents on years of membership in Farm Bureau. The
mean years of membership for mail respondents was 23.4 years, with a standard deviation of
14.3. The mean years of membership for phone respondents was 17. 1 years, with a standard
deviation of 11.8. It is possible that this difference was due to measurement error, since years of
membership for mail respondents was collected from the mailed questionnaire, whereas years of
membership for phone respondents was collected from FFBF membership records. FFBF staff
members recognize that the membership database is likely to under represent true years of
membership, since it is based on consecutive years paid. (As previously mentioned, if a member
was late on paying annual membership dues, the count of consecutive years of membership
would start back at one.)
The difference between mail and phone respondents could be attributed to measurement
error and survey respondents considered representative of non-respondents. However, ANOVA
tests on ratings of performance for select best practices also revealed significant difference
between mail respondents and phone respondents. The most extreme difference was in board
performance ratings on the best practice of "share Farm Bureau facts and history in public
relations activities." The mean rating for the mail respondents was 2.20, with a standard
deviation of 0.75, while the mean rating for phone respondents was 1.67, with a standard
deviation of 0.76. Analysis of variance revealed a significant difference, F(1, 412) = 14. 16,
p<.000. Statistical differences were observed with four other practices, but those differences
were less extreme (i.e. p>.001).
Such differences between groups could be attributed to differences in survey mode (self-
administered mail questionnaire vs. telephone interview). Dillman (2000) and others have
recognized consequences of mixed-mode designs. Several researchers have noted differences in
answers provided over the telephone compared to answers provided in a self-administered mail
questionnaire (Dillman, 2000; Schwarz, Strack, Hippler, & Bishop, 1991). The results of the
tests comparing mail respondents and phone respondents casts some doubt on assumed
similarities between respondents and non-respondents. Further investigation is warranted.
In order to further investigate possible differences between respondents and non-
respondents, the researcher compared early respondents to late respondents. The method is
based on research that shows late respondents are often similar to non-respondents (Ary et al.,
2002; Clausen & Ford, 1947; Lindner et al., 2001; Miller & Smith, 1983). For this study, early
and later respondents were classified by waves of response, a strategy that has been supported by
other researchers (Israel, 1992; Lindner et al., 2001). As questionnaires were mailed, each was
given a number, and this number was used to identify each "wave." The second wave of
questionnaires mailed to non-respondents was numbered beginning with a "2" (either 200 or
2000, depending in which instrument). By distinguishing early and late respondents based on the
questionnaire number, evidence suggests that late respondents were not only late respondents,
but they were further persuaded to respond by the additional attempt to reach them. In fact,
many may have discarded the first questionnaire but responded to the second questionnaire in
order to avoid the nuisance of additional contacts. The early-late analysis compared 298 early
respondents with 113 late respondents. Tests for ANOVA revealed no significant difference on
the following demographics: age, gender, level of education, year of membership in FFBF, years
of service on the local FFBF board, and number of local FFBF board meetings attended in the
last six months. In addition, ANOVA test on ratings of performance for select best practices
revealed no significant difference between early respondents and late respondents. The results of
these tests suggest that no significant difference exists between study respondents and non-
Although the early to late comparison method suggest that no significant difference exists
between respondents and non-respondents, caution should still be exercised in generalizing
beyond the sample. In their discussion of the early to late comparison method, Clausen and Ford
(1947) concede that "the estimate for non-respondents is clearly in the nature of a guess, but it is
an informed guess" (p. 507). Some doubt remains, particularly since phone respondents, as a
representation of non-respondents, showed some significant differences from mail respondents.
The difference may be a function of the difference in survey mode (self-administered mail
questionnaire vs. telephone interview) (Dillman, 2000; Schwarz et al., 1991). The caution
required for extending findings from the study to the larger population may depend on the
severity of consequences that might result in doing so. In other words, the data from this study
can be assumed to be representative of the larger population provided that doing so is not likely
to have negative effects on the population.
Early in the data collection period, the researcher noticed item non-response among
multiple returned questionnaires. Item non-response can prevent observations from being
included in the data analysis, which in turn reduces statistical power (Gravetter & Wallnau,
2000) and can introduce bias in the data (Israel, 1992; Miller & Smith, 1983; Ognibene, 1971).
Among the returned questionnaires in this study, 269 (65.6%) were complete, while 141
(34.4%) had missing data for one or more variables. Of the questionnaires with missing data, 13
were considered extreme in that they failed to answer more than half of the questions. The
missing data from these observations would have limited the practical value of including them in
the data analysis. As a result, the 13 observations with extreme missing data were treated as unit
non-response and dropped from additional analysis. The resulting sampling population included
397 study participants, 81 from the treatment group, 76 from the control group, and 240 from the
spillover treatment group. Item non-response by group is outlined in Table 3-2. A final usable
response rate for each group is identified in Table 3-3.
Dependent Variable Indexes
Individual-level dependent variable
As previously stated, the dependent variable for this study was performance on Farm
Bureau related practices. On an individual level, this performance was measured by the survey
instrument as a self-rating. This variable was created as an index of the individual ratings of the
17 best practices identified in the survey instrument. Study participants were asked to indicate to
what extent they implemented each practice during the preceding six months. Response options
included: not at all, somewhat, mostly, and completely. These responses were coded "O", "1",
"2", and "3", respectively. The researcher used these ratings to form an index for overall
individual performance on the Farm Bureau practices. In order to minimize the effects of
missing data, the index score was calculated as an average of the ratings provided by the
respondent. For example, if a respondent marked "somewhat" for eight practices and "mostly"
for eight other practices but failed to provide a response to the remaining practice, the
respondent' s average rating of 1.5 would be kept in the analysis, virtually unaffected by the
missing value. If, however, a respondent failed to provide ratings for at least half of the
practices, that respondent' s index score was coded as missing and not included in further
analysis. This was done to prevent excessive missing data from providing added weight to the
responses that were provided. The researcher evaluated the final index for reliability and found a
Cronbach's alpha (a) of 0.89. According to Traub (1994), indexes in the social sciences that
have a reliability coefficient of 0.8 or larger are often considered well-constructed and relatively
precise. As a result, this index was considered reliable.
Although self-assessments of performance deserve some skepticism for their potential bias,
they have been widely used in research with nonprofit organizations. Preston and Brown (2004)
put it simply when they said, "to measure individual performance effectively is a challenge" (p.
227). This challenge is further complicated in organizations like Farm Bureau, that rely heavily
on volunteer performance without regular supervision. In a study that investigates short-term
changes in personal performance, as was the case for this study, few options exist for obj ective
evaluation, and it may be that the most valid evaluation is a self-assessment.
Board-level dependent variable
Board level performance was also measured in the survey instruments. As a dependent
variable, board level performance was calculated as an aggregate of individual board members'
average assessment of their board' s implementation of the 17 best practices identified in the
questionnaire. The individual average rating of the board was calculated in a similar manner as
was done for each individual's self-performance rating. If a respondent failed to provide board
ratings for at least half of the practices, that respondent' s index score for board performance was
coded as missing and not included in further analysis. Once individual ratings of board
performance were calculated, those ratings were sorted and averaged by the local board (i.e.
county) associated with the rating. If a local board had an individual board rating from only one
board member, that board's performance rating was equated to the one available rating.
However, most local boards had multiple respondents, so the board's performance rating was an
average of several board members' assessments of board performance among the 17 practices.
The researcher evaluated the index for reliability and found a Cronbach' s alpha (a) of 0.91. As a
result, the index was considered reliable.
Local board performance among the 17 best practices was also evaluated by FFBF staff
through the "Field Staff Perceptions Survey." The field staff members' assessments of local
board practices were averaged for each county/local board to create an index of overall board
performance. The researcher evaluated the index for reliability and found a Cronbach's alpha (a)
of 0.93. As a result, the index was considered reliable. In addition, the researcher investigated
the relationship between the board performance rating by the field staff and the board
performance aggregate rating from the local board members. The researcher found a correlation
of .343, with a p-value of .026. This statistic suggests a moderate relationship between the two
ratings (Cohen, 1988, pp. 82-83; Penfield, 2003, p. 185). The relationship is important to
consider, because the data analysis in this study relies heavily upon subj ective ratings and the
field staff ratings represent the only data available as an obj ective evaluation that provides a
comparison across local/county boards. One of the reasons why the FFBF field staff member
ratings were not included in final regression models as a dependent variable is because the
ratings were not provided for 18 of the 61 counties, thus limiting the analysis of data available
from those counties. Instead, the local board members' aggregate rating of board performance
was the dependent variable associated with board-level performance.
Once again, concerns about potential bias from self-assessment should be considered.
However, the differences in performance ratings between board members and FFBF field staff
do not necessarily suggest problems with validity. Local board members and field staff members
can be considered different stakeholder groups, and previous research with nonprofit boards has
shown that "stakeholders do not judge organizational effectiveness similarly" (Herman & Renz,
2004, p. 699). In a study of 44 nonprofit organizations, Herman and Renz found that, "for board
members, the use of correct management practices is strongly and positively correlated to
organizational effectiveness judgments" (p. 700), whereas the same situation failed to have a
significant impact on organizational effectiveness judgments by funders and senior staff (p. 701).
Another reason for the differences in performance ratings between board members and Hield staff
is that "nonprofit organizational effectiveness is always a matter of comparison" (Herman &
Renz, 2002, p. 3). This comparison is likely to occur even when attempts are made to objectify
the measures. Because of their knowledge of other local boards, the Hield staff members are
likely to rate individual board performance on the local Farm Bureau practices as compared to
other local boards. Board members, on the other hand, are likely to rate individual board
performance on the local Farm Bureau practices as compared to earlier observations.
Several nonprofit researchers have worked to develop a standardized self-assessment
instrument, the Board Self-Assessment Questionnaire (BSAQ), to evaluate board performance
across organizations (Holland, 1991; Jackson & Holland, 1998). The BSAQ has been used with
more than 100 nonprofit boards and has shown strong validity when examined by knowledgeable
raters of board and organizational performance. The BSAQ was not appropriate for this study
with FFBF local boards because it was not designed and validated for use with smaller nonprofit
organizations that are staffed mainly by volunteers.
Part of the reason why empirical evidence on nonprofit board performance relies heavily
on self-assessment may be due to the belief that "nonprofit organizational effectiveness is a
social construction" (Herman & Renz, 2002, p. 4). "How to conceive and measure
organizational effectiveness is an issue about which there is no scholarly consensus" (Herman &
Renz, 1997, p. 13). Instead, researchers rely on stakeholders to judge organizational
effectiveness. In the case of grassroots organizations, like Farm Bureau, the local leaders are key
stakeholders and therefore should be a primary resource for evaluating effectiveness.
Independent Variable Indexes
As previously mentioned, independent variables analyzed in this study included constructs
associated with goal theory. These constructs were measured in the "Best Practices Goal
Survey," administered to .So eigthenl~lila the Voice program participants. The section titled, "Best
Practices Goal Setting Process..." included 24 goal-related questions, all measured on a Likert-
type scale that included: "strongly disagree," "disagree," "neither agree nor disagree," "agree,"
and "strongly agree." When item nonresponse was observed, the researcher left the nonresponse
out of the analysis for the individual items. However, after the analysis of descriptive statistics,
the approach to nonresponse changed. For the purpose of keeping observations in the goal
constructs, the researcher coded item nonresponse as "neither agree nor disagree." Although this
practice increases the potential for measurement error, the researcher assumed that such effects
would be minimal and that the benefits of recoding the data would outweigh the consequences.
An ideal research study of goal setting theory would measure all 12 constructs suggested
by the theory and the high performance cycle. However, measuring all 12 constructs exceeded
the limitations of the instrument and the sample size. The instrument did measure two key
components of the goals: content and intensity. The goal content was expressed in two aspects:
specificity and difficulty. Goal intensity was measured through goal commitment and self-
efficacy. An addition, the instrument measured a construct for perceived goal performance.
Once the data were collected, the researcher conducted factor analyses to analyze and
refine the item components of each construct. Tests of reliability were conducted using
Cronbach's alpha. When considered as a group, the estimated reliability of all goal-related
questions was .877. For the individual constructs, specificity had a reliability of .738, difficulty
had a reliability of .763, commitment had a reliability of .838, and self-efficacy had a reliability
of .708. The goal performance construct had a reliability of .889.
According to Penfield (2001, cited in Moore, 2003, p. 65), reliability coefficients that
exceed .90 are considered high, those that exceed .80 are considered moderate to high, and those
between .70 and .80 are considered low. With this classification, the overall goal ratings had a
moderate to high reliability, while some of the constructs had low reliability. However, as Traub
(1994) points out, the primary standard for judging acceptable reliability is the reliability of
available alternative measures. With this in mind, the researcher continued analysis with the less
reliable constructs, recognizing that measures of improved reliability should be sought in future
research. The final reliability and factor loadings of each construct are displayed in Tables 3-4
The constructs measured in this study were arranged in a hypothesized path model.
Consistent with goal setting theory, the researcher hypothesized that the goal core (specificity
and difficulty) would have a positive impact on performance (Figure 3-3). In addition, the
researcher hypothesized that goal specificity would positively impact goal commitment and self-
efficacy (as related to the goal), while goal difficulty would negatively impact self-efficacy.
Self-efficacy, in-turn, was expected to positively impact goal commitment, and goal commitment
was expected to positively impact goal performance. Although goal mechanisms were not
measured in this study, goal setting theory would suggest that goal mechanisms also mediate
these impacts on performance.
Farm Bureau Foundations program participation
Individual attendance at the FarmFFFFF~~~~~~~FFFFFF Bureau Foundations workshops was recorded by FFBF
and either validated or corrected by survey respondents. The number of participants from each
local board was then investigated for various levels of participation. The most basic level of
participation was a simple count of local board attendees. However, in order to investigate
possible non-linear relationships, this count was also dummy coded for participation levels of
one or more, two or more, three or more, four or more, five or more, and six or more. In
addition, the percent of local board members attending the workshops was calculated by dividing
the number of local board member attendees by the total number of directors for that local board.
In order to investigate for non-linear relationships, this percentage was also dummy-coded for
participation levels of 10%, 25%, 33%, and 50%. The raw numbers and dummy codes were both
analyzed when investigating the effects of workshop attendance by a local board.
Weighted goal theory constructs
In order to include goal theory constructs at a local board unit of analysis, the ratings on
these constructs needed to be considered across all directors in a given local board. However,
since not all board members participated in the goal setting process and therefore did not have
ratings for the goal constructs, the available ratings needed to be weighted in a way that
considered level of participation. In order to accomplish this, the aggregate ratings for a local
board were multiplied by the percentage of board members who participated in the FarmFFFFF~~~~~~~FFFFFF Bureau
Foundations workshops, and thus were engaged in the goal setting exercise. Because the goal
construct ratings were centered on zero, with a range of negative two to positive two, this
weighting system equated non-participation as rating of zero (neither agree nor disagree) on the
While this weighted rating was a synthetic representation of actual individual ratings, the
researcher believed it to be the best approach for considering the goal constructs at a local board
unit of analysis. If an individual had responded "neither agree nor disagree" for all goal-related
variables, the information would have no influence on the aggregate board rating for the goal
constructs. In another instance, if an individual had responded "strongly agree" for all of the
goal-related variables, but that board member was the only director engaged in the process, the
goal construct rating would be attenuated by the lack of participation among the entire board.
On the other hand, if the entire board participated in the FarmFFFFF~~~~~~~FFFFFF Bureau Foundations program
component and had similar ratings about their experience with the best practices goals, then this
fact would carry more weight in the analysis. This method of weighting the ratings is somewhat
experimental and should be considered when interpreting the data.
Few of the board demographics considered in this study were directly measured. The total
number of local board members was derived from the FFBF field services directory. However,
data on other board-level independent variables were calculated from information provided by
survey respondents. For local boards that had high response rates in this study, the calculated
demographics are highly valid estimates of the local board's true demographics. However, for
local boards with lower response rates the calculated demographics may be less representative.
An extreme example is a local board that had one survey respondent, and the board member was
female. It is highly unlikely that the local board is comprised entirely of female directors, so the
estimate of 100% female board members likely contains measurement error. However, the
calculated demographics were the best available source for this information. In addition,
because the board-level dependent variable was based entirely upon ratings from survey
respondents, the observed relationship between the board demographics (independent variables)
and the rated board performance (dependent variable) were expected to be similar to what would
be observed for the true population. Even still, some caution should be exercised in interpreting
the effects of board demographics analyzed in this study.
Analysis for Objectives One Through Three
Descriptive statistics, including frequencies and measures of central tendency, were used to
organize and summarize the individual item response data from the collected questionnaires.
Findings for obj ectives one through three were obtained primarily through the use of descriptive
statistics. When appropriate, analysis of variance (ANOVA) was used to investigate differences
Analysis for Objectives Four and Five
The goal-related hypotheses to be tested in obj ectives four and five are as follows:
1. Units with higher perceived levels of goal specificity will also have higher levels of
2. Units with higher perceived levels of goal difficulty will also have higher levels of
3. Units with higher perceived levels of goal specificity will also have higher levels of goal-
4. Units with higher perceived levels of goal specificity will also have higher levels of goal
5. Units with higher perceived levels of goal difficulty will have lower levels of goal-related
6. Units with higher perceived levels of goal-related self-efficacy will also have higher levels of
7. Units with higher perceived levels of goal commitment will also have higher levels of
These hypotheses are graphically displayed in Figure 3-3.
Correlations between the independent and dependent variables were analyzed to examine
the effect of the independent variables on the dependent variables on an individual basis. The
magnitudes of the correlations were described using terms and classification appropriate for the
context of social science research. A correlation of 0. 10 is described as a small effect size (weak
relationship), a correlation of 0.30 is described as a medium effect size (moderate relationship),
and a correlation of 0.50 is described as a large effect size (strong relationship) (Cohen, 1988, pp.
82-83; Penfield, 2003, p. 185).
In an attempt to explain the influence of demographic and program variables on best
practices goal achievement, the researcher used multiple regression to build explanatory models.
According to Gliem (2003), multiple regression is "a method of analyzing the variance of a
dependent variable (Y) by using information available on two or more independent variables" (p.
1). The basic model may be expressed as:
E(Y) = a + P1X1 + P2X2 PkXk
In this equation, a, pr,..., Pk are parameters (properties of the population) and X1, X2,... Xk
are explanatory (independent) variables, where k denotes the number of predictors. Y is the
response (or dependent) variable, and E indicates that it is an estimate or prediction. In this
study, the Y variable is the best practices adoption score. Important assumptions associated with
the use of multiple regression include: independence, linearity, normality, and equal variances
(Algina, 2005, p. 391). An alpha level of 0.05 was set a priori for the statistical analysis.
However, varying levels of alpha were considered and reported because the knowledge of these
variations improved the exploratory value of the study.
One practical issue to consider in the implementation of multiple regression analysis is the
ratio of cases to independent variables. "Numerous rules-of-thumb have been suggested for
determining the minimum number of subj ects required to conduct multiple regression analyses"
(S. B. Green, 1991, p. 499). Green has compared these methods and argues that the rules-of-
thumb that specify some constant, such as a minimum number of subj ects or a minimum ratio of
subjects to predictors, are not sufficient because they do not consider effect size and its
contribution to statistical power. Green recommends a more complex rule-of-thumb that
"estimates minimum sample size as a function of effect size as well as the number of predictors"
(p. 499). However, Green's recommendation is based on a power analytic framework that is
used to ensure a reasonable chance of rej ecting null hypotheses involving regression parameters.
Since this is an exploratory study that was a census of the population, the researcher did not
apply Green's more complex recommendation. Instead, the researcher chose to follow a more
generous standard, proposed by Tabachnick and Fidell (1989), setting the minimum number of
cases or subj ects for each independent variable at five to one.
The usefulness of regression models is evaluated by the coefficient of determination,
denoted by R-Square (R2). The coefficient of determination represents "the proportion of
information in the dependent variable that is explained or accounted for by the independent
variable" (Penfield, 2003, p. 231). In the context of social science, researchers have provided
general "rules of thumb" for interpreting the value of R2. An R2 Of 0.01 represents a weak
relationship (small effect size), an R2 Of 0.09 represents a moderate relationship (medium effect
size), and an R2 Of 0.25 represents a strong relationship (large effect size) (Cohen, 1988, pp. 79-
81; Penfield, 2003, p. 232). This classification was applied in interpreting the coefficient of
determination (R2) in this study. Adjusted R-Square values, rather than the raw coefficients, will
be reported because they offer a less biased estimate for analyses with a small number of
observations and numerous independent variables (Agresti & Finlay, 1997; Tabachnick & Fidell,