Review5

Journal of Sport Sfanagement, 1990. 4. 161-173

Athletic Director Leader Behavior as a Predictor

of Intercollegiate Athletic Organizational Effectiveness

Dallas Branch Jr. West Virginia University

Intercollegiate athletics has come under increasing scrutiny. Questions of leadership and the NCAA's Presidents' Commission reflect new levels of exposure and commitment to clean the athletic house. The problem of defining the academic/athletic balance in big-time college sports has polarized faculty, administrators, and athletic leaders at many cdleges and universities. The purpose of this study was to examine athletic director and selected assistant perceptions of leader behavior to detennine whether their perceptions con- tributed significantly to the prediction of imercollegiate organizational effec- tiveness. Findings indicate that effective athletic organizations have leaders who are more predisposed to goal and task accomplishment than to developing good interpersonal relationships with their subordinates. Contemporary leader- ship theory and management philosophy suggests that organizations that can accomplish both are most effective. Athletic directors may want to adjust their leadership behaviors to meet the managerial demands of today's inter- collegiate athletic program.

In America today, the responsibilities and leadership roles of Division I-A intercollegiate athletic directors are undoubtedly changing. In the past, ADs were selected primarily for their playing or coaching skills and were assumed to pos- sess the requisite knowledge and skill level to perform in largely public relations capacities (McGee, 1984). Many had had stellar careers as players and/or coaches in college or professional sports (Landry, 1983). Few received formal educa- tional preparation to assume their roles as leaders of increasingly conqjlex busi- ness organizations. Most were schooled in the education or physical education disciplines instead of in business or sport management (McGee, 1984). Hatfield, Wrenn, and Bretting (1987) have substantiated this earlier finding, as 36 and 37% of the 105 Divisioi I-A athletic directors received their undergraduate and master's

Dallas Branch Jr. is with the Department of Sport & Exercise Studies, West Virginia University, Morgantown, WV 26506.

161

162 Branch

Deflnition of Terms

Athletic director—The individual responsible for planning, organizing, leading, and evaluating a program of intercollegiate athletics, A Division I-A athletic director operates a program designated by the National Collegiate Athletic Association (NCAA) as one of the 105 athletic programs competing in the top division in football.

Associate/assistant athletic director—A subordinate to the athletic director in tbe administrative hierarchy who assists in managing, directing, and administering an inter- collegiate athl^c program. For purposes of this study, he/she is a randomly selected subor- dinate to the Division I-A athletic director responsible for supporting administrative responsibilities in a Division 1-A intercollegiate athletic program. Often this individual has specific intradepaitmental responsibilities in a particular area of operation such as athletic business (budgeting), marketing, promotion, sports information (publicity), or fund-raising.

Perceived leadership behavior—The manner of behaving or acting reflected in the role of the group leader, in this study the professional behavior of the athletic directors as perceived by themselves and by selected associate/assistant athletic directors. In past studies, the evaluative dimensions of consideration and initiating structure, as measured by the Leader Behavior Description Questionnaire-Form XH (LBDQ-Form XII), are combined above and below the split mean on each dimension to distinguish between high and low perceptions of behavior (Stogdill, 1969; Stogdill, Goode, & Day, 1962).

Initiating structure—Refers to the leader's behavior in "delineating the relationship between himsdf and the members of his group, and in endeavoring to establi^ well-defined pattems of organization, channels of communication, and ways of getting the job done" (Halpin, 1957, p, 1), For the purposes of this study, initiating structure represents the extent to which the athletic director, as a leader of the athletic organization, is perceived to organize the relationship between himself and subordinate members of the athletic pro- gram, to define the various role expectations of the subordinates, and to establish well- defmed pattems of organization and channels of communication to accomplish the goals and objectives of the athletic program (Austin, 1975).

Consideration—Refers to the "behavior indicative of friendship, mutual trust, respect, and warmth in a relationship between the leader and members of the group" (Halpin, 1957, p. 1), For the purposes of this study, it refers to the behavior of athletic directors as perceived by themselves and selected subordinates on how well they work together in developing friendship, mutual trust and respect, and warmth in the course of their day-to-day interactions to promote better interpersonal relations.

Athletic organizational effectiveness—Defined as the integration of five effective- ness criteria by Mott (Hoy & Miskel, 1982) which include quality and quantity of the organization's "product," its efficiency, its ability to adapt to changes, and its flexibility in dealing with crises. Adapted for use in intercollegiate athletic settings, the Perceived Index of Athletic Organizational Effectiveness measures the ability of organizations to mobilize their centers of power to achieve goals and to a d ^ to environmental changes (Mott, 1972). For the purposes of this study, athletic organizational effectiveness is de- fined as the ability of Division I-A intercollegiate athletic organizations to be productive (quality and quantity of the work), to be efficient in their (^rations, to adapt readily to changes prescribed by intemal and/or external constituencies (public perceptions), and to be flexible in handling unexpected crises (NCAA investigation, scandal, etc.).

Athletic Director Leader Behavior nxJ

degrees in physical education, respectively. Though well-intentioned, these ADs managed their athletic departments as they coached or had been coached them- selves: their word was law (Austin, 1975). This "trial and error method has frequently been the most effective teacher'' in getting the job done in college athletics (McGee, 1984, p. 20). These phenomena may have given rise to the "good ole' boy" network in intercollegiate athletics that seems to prevail even today.

Not that athletic coaching or playing experience doesn't account for some- thing worthwhile in the profession—quite the contrary. One should have an under- standing of the unique environment within a collegiate athletic program. Leaders in this arena should also maintain a level of empathy for student-athletes and their commitment, sacrifice, and dedication to athletics. Those former players and coaches now occupying athletic director positions app)ear to have these requisite qualities.

Intercollegiate athletics has become an increasingly visible arena in higher education. Intensified marketing of college athletics, greater involvement of col- lege presidents in the governance of college sports, and frequent charges of impropriety are tnajor departures from past practice which reflect new levels of exposure. Recent issues of Sports inc.. Sports Illustrated, and Time have reported on problems confronting big-time college athletics. At any given time one can find some 25 Division I-A athletic programs under NCAA sanctions ("25 Institutions," February 15, 1989).

The problems associated with the big business of intercollegiate sports suggest a need for more effective leadership at a time when there appears to be a leader- ship vacuum. Krupa and Dunnavant (1989) address the downside of cdlege sports: "Whether intercollegiate athletics officials seize the initiative and fashion their own systems, or whether changes occur in a leadership vacuum, the face of college athletics will appear far differently by the year 2000" (p. 33). A preoccupation with winning and the fiscal bottom line has sullied the integrity of college sports by ' 'continuing to perplex a public that expects more idealistic behavior from institutions of higher learning" (Krupa & Dunnavant, 1989, p. 37).

Increased scrutiny of intercollegiate athletics may have forced big time col- lege athletic directors to examine their roles and assess their leadership abilities. As a result, some ADs are beginning to gain the respect, prestige, and responsi- bility that make them educational and athletic administrators (Dulac, 1989). Under- standing this new leadership role as it relates to organizational tasks and to interpersonal concerns of futiu« athletic organizations may help direct and focus the attention of future athletic directors. The risk of jeopardizing institutional and public support for college sports suggests a need for tjo less than appropriate skills and efforts firom leaders of intercoUegiate athletic programs.

Responsibility for the integrity problems confronting coUege athletics can be shared by all members within the structural framework of the respective institutions. However, the locus of power, where ultimate decisions are made, clearly should rest with each organization's leader—the athletic director (Hoy & Miskel, 1982). One should not discount the influence of close associates and the impact of decisions made at higher levels in the institution, as they largely determine the appropriate role of athletic leaders and the infiuence they are able to exert (Chelladurai, 1985). Thus, a clarification of the nature of effective ath- letic leadership may be of value to a variety of persons within and extemal to every intercollegiate athletic organization.

164 Branch

Statement of the Problem. The question asked in this study was whrther selected dimensions of athletic leader- ship behavior, as perceived by the 105 Division I-A major college athletic direc- tors themselves and by a randomly selected associate or assistant athletic director from each institution, would contribute significantly to the prediction of athletic organizational effectiveness. Specifically, the study sought to detennine the rela- tive contributions of "consideration" and "initiating structure," as fundamental leadership behavior constructs, to the prediction of intercollegiate athletic organiza- tional effectiveness.

Purposes of the Study

There were two underlying purptoses of this study. One purpose was to gain empiri- cal evidence from which to promote a better understanding of collegiate athletic administrator leadership behavior as it relates to the effectiveness of the athletic organization. Intercollegiate athletics has largely been ignored as an arena for empirical investigation, research, and study. This statement would be seconded by McGee (1984), who quotes Richard Block, principal owner of the Phoenix Suns, that "college athletics is one of the few businesses that no one knows any- thing about" (p. 20). Athletic directors need more knowledge of themselves, their leadership roles, and their interactions with subordinates (Aceto & Ryckman, 1972). This study attempted to shed some light on the arena of intercollegiate athletics, which has been relatively ignored as an area of research and inquiry. The second purpose of this study was to test a multiple regression model for its goodness-of-fit to the study's data to determine whether a selected set of leader behavior predictors and a dependent measure of athletic organizational effective- ness were linearly related.

This study did not intend to resolve the leadership controversy between a one-best leadership orientation and a flexible leader behavior that changes to meet situadonal demands. The controversies between these two orientations are well articulated in the leadership literature (Blake & Mouton, 1964; Fiedler, 1967; Hershey & Blanchard, 1977; Hoy & Miskel, 1982; StogdiU, 1948).

Methodology

Instrumentation and Measurement Scales

Three questionnaires were used to coUect the data in this study. The Perceived Athletic Organizational Effectiveness Index, an adapted organizational effective- ness instrument developed by Mott (1972) for use in educational settings, was modified for the purposes of this study in adiletic organizations. It contained eight questions designed to measure the effectiveness of die attiletic organization as per- ceived by the athletic director. A Likert scale evaluated each response on a 1 to 5 basis. The organization's effectiveness score, ranging from 1.00 to 5.00, was derived by dividing the total score (8 to 40) by the number of questions (8). A score of 1.00 represents an ineffective athletic organization while a score of 5.00 represents an effective athletic organization. The index's reliability is high (0.89) (Hoy & Miskel, 1982). Independent raters assessed the internal consistency of the instnunent. Rank-order correlations of between .72 and .55 suggest "considerable

Athletic Director Leader Behavior 165

agreement on the relative effectiveness of the various dimensions" (Mott, 1972, p. 193). Barcikowski (personal cotnmunication, June 1985) reviewed minor modi- fications to the instrument, primarily wording to refiect appropriate organizational context, and concluded that the changes wotild not alter its internal consistency or reliability.

The second and third questionnaires represented versions of the LBDQ- Form Xn that were mailed to athletic directors and selected associate/assistant athletic directors. The LBDQ-Form XII consists of 100 questions representing 12 dimensions of leadership behavior. For purposes of this study, consideration and initiating structure dimensions were extracted for analysis. Each dimension is represented by a series of 10 questions. The subjects indicate their response by circling one of five letters (A,B,C,D,E) that correspond to five scores (5,4,3,2,1). The range of scores on each of the two dimensions is between 10 and 50. Szilagyi and Keller (1976) report test-retest reliability coefficients of .89 for the consideration subscale and .87 for the initiating structure subscale. In addi- tion, they report congruency coefficients between .80 and .99 for the leader behavior dimensions studied, which indicate strong cross-sample factor stability. Greene reports test-retest reliability of between .57 and .72 on the initiating struc- ture subscale and between .71 and .79 for the consideration subscale, as determined by a modified Kuder-Richardson formula (Buros, 1978).

Subjects

The population of 105 Division I-A athletic directors and programs were selected to participate in this study to provide a self-perception of their leadership behavior. Division I-A athletic directors were selected for this study because their athletic programs represent the highest competitive level of play in college football. In addition. Division I-A athletic programs are primarily responsible for the perceived lack of integrity in college sports in America today.

In order to assess a subordinate perception of athletic director leadership behavior, an associate/assistant athletic director was randomly selected from those in each program to assess their perceptions of the athletic director's leader behavior. Each Division I-A athletic program has at least one associate and/or assistant athletic director. Each program's subordinate group was sampled as a referent "other" perception of athletic director leadership behavior.

There were three purp)oses of the subordinate selection process. First, the associate/assistant works closely with the athletic director on a daily basis and could provide a more objective percqjtion of the director's leadership behavior. Second, the associate/assistant was selected on the basis of general departmental responsi- bilities rather than specific areas of expertise, thus minimizing "maturity level" and "proximity to the leader" effects which are stated limitations by Hershey and Blanchard (1977) and Mott (1972). Finally, the perceptions of associate/ assistant directors could be matched with those of the athletic director from the same program.

Data Collection

Barcikowski (1980) outlines a systemic procedure for data collection using mailed questionnaires. Stogdill (1%3) also provides useful insight into administering the LBDQ-Form Xn. The following procedure was used in this study: (a) A list was compiled of all Division I-A athletic programs, (b) One associate or assistant

166 Branch

athletic director from each program was randomly selected, (c) Endorsements for the study were solicited from a recognized member of each ^h l^c conference, the National Association of Collegiate Directors of Athletics (NACDA), and the coordinator of Ohio University's Sports Administration and Facility Management Program. These endorsements appeared on all cover letters accompanying the questionnaires, (d) The Perceived Athletic Organizational Effectiveness Index questionnaire was mailed to all Division I-A athletic directors for them to evaluate ttieir athletic program's effectiveness, (e) Two weeks later the LBDQ-Form XII questionnaires were mailed to all athletic directors for their perceptions of their own leader behavior., and to all selected associate/assistants for their perceptions of their athletic director's leader behavior, (f) A cover letter accompanied each of the three questioimaires mailed, (g) Two weeks later, nonrespondents were sent foDow-up letters and questionnaires, (h) If responses were not received within 3 weeks, the nonrespondents were telephoned and were asked for responses to the 20 leader behavior questions pertaining to consideration aixl initiating structure.

Of the 105 effectiveness questionnaires distributed, 88 (84%) were returned. In addition, appropriate versions of the LBDQ-Form XII were sent to each ath- letic director and the randomly selected associate/assistant athletic director (subor- dinate) from each athletic program. Of the 105 questionnaires sent to each group, 78 (74%) and 87 (83%) were utilized, respectively (see Table 1).

Statistical Design

A hierarchical multiple-regression analysis was used to examine the relationship between the dependent (criterion) variable, athletic organizational effectiveness, and the indepraident (predictor) variables (see below). TTiis analysis permits each predictor variable to be entered into the regression model in a logical manner. The ordering of predictor variables was established a priori based on previous leadership research and theory. This allowed the contribution of each predictor variable to be assessed while controlling for the influences of previously entered

Table 1

Questionnaire Returns

Questionnaire

Athletic org. effective- ness

LBDQ-subordinate LBDQ-athletic director

Mailed n

105 105 105

Returned n

90 81 71

86 77 67

Refused n

2 3 2

Useable n

88 78 69

%

84 74 65

Returned by phone*

n

9 9

Total n

88 87 78

%

84 83 74

'Note. Hotelling t tests were performed on the data to determtae if significam differences in responses were found in the mode of data collection (mail vs. phone). Norte were found, thus responses gathered in both modes were pooled

Athletic Director Leader Behavior 167

variables. Each predictor was analyzed for its individual contributions to the prediction of athletic organizational effectiveness. In addition, the interactions of joint contributors were assessed to analyze the combined effects of two predic- tors while controlling for the effects of individual and/or other joint predictors.

Variables that the literature suggested would be the best predictors of organizational effectiveness, that is, consideration and initiating structure as perceived by the athletic directors themselves, were entered individually into the regression equation in the initial two steps. In Steps 3 and 4 each of these two dimensions, as perceived by selected associate/assistant athletic directors, was entered individually to assess its relative contributions to the predictive power of the regression model, controlling for the effects of previously entered contri- butors. Two interaction variables, products of consideration and initiating structure as perceived by the athletic directors and tlwir subordinates, were entered into the regression model in Steps 5 and 6 while controlling for the effects of previously entered individual contributors.

Raw score regression equation:

Y = a -I- b,x, + bjXi + bsXa -I- b4X4 + bsXj -)- bex* -I- e

Standard score regression equation:

Zy = /3,z, -t- /SiZj -I- /SsZj

where Y = raw score—athletic organizational effectiveness; Zy = standard score—athletic organizational effectiveness; a = constant term; e = error term; Xj = consideration as perceived by athletic directors; X2 = initiating structure as perceived by athletic directors; X3 = consideration as perceived by subordinates; X4 = initiating structure as perceived by subordinates; Xs = interaction of x, and X2; Xt = interacticm of X3 and X4; b,-b6 = raw score regression coefficients; iSi-^e = standard score regression coefficients; Zi-z« = standard scores corresponding to variables Xi-x«.

The SPSS (X) regression statistical program permitted a logical ordering of leader behavior dimensions in the regression model based on available litera- ture to detennine the individual and joint contributions made to the prediction of athletic organizational effectiveness (Nonisis, 1983). The statistical test used to detennine the significance of contribution to the regression model by each dimension singly and by the interaction of dimensions jointly was the F test. The criterion for determining the predictive power of the model at each of the order- ing steps was the significant change in the iP statistic. Each of the six questions of interest was tested at a level of significance established a priori at .01 to con- trol for the experimental error.

The r^ression equation represents a mathematical model that relates ath- letic organizational effectiveness linearly to a specific set of leader bdiavior dimensions to determine whether any or all of these dimensions are important predictors of effectiveness. The ideal model is one in which the set of predictor variables is highly correlated with the dependent variable (athletic organizational effectiveness) but is relatively nonintercorrelated, thus representing a goodness- of-fit. These conditions allow for different portions of the variance in the crite- rion (dependent) variable to be accounted for by each predictor while controlling for the influence of oth» predictors.

168 Branch

Table 2

DescripUve Statistics

Variable n M SD Min. Max.

Consideration as perceived by attiletic director (X CON) 77 38.87 4.08 29.00 48.00

Initiating structure as perceived by athletic director (Y IS) 77 40.61 4.00 28.00 49.00

Consideration as perceived by subordinates (X3CON) 78 35.45 7.41 16.00 50.00 Initiating structure as perceived by subordinates

(X^IS) 78 38.67 5.83 26.00 50.00 Athletic organizational effectiveness (YAOE) 84 3.74 .45 2.50 4.88

Summary of Findings Descriptive statistics are provided in Table 2 for the individual leadership behav- ior dimensions (predictors) and athletic organizational effectiveness (criterion) measure examined in this study. Initiating structure, as perceived by the athletic directors (X2IS), was rated highest among the four predictors (M=40.61). The next highest rating was for consideration, as perceived by athletic directors (M=38.87). Subordinate perceptions (mean scores of X3CON and X4IS) of ath- letic director leadership behavior were lower (35.45 and 38.67, respectively) than those given by the athletic directors themselves, indicating that ADs perceived their behavior more favorably than did their subordinates. Also, higher standard deviations of subordinate percq)dons (7.41 and 5.83) versus those of athletic direc- tors (4.08 and 4.00) indicate a wider disparity between AD and subordinate per- ceptions, although the congruency of leadership behavior perceptions was beyond the scope of this study.

In addition to understanding the relative strength of a set of predictors, multiple-regression analysis permits building a model with one or more predic- tors to test its goodness-of-fit (predictive power). The ideal set of predictors is one highly correlated with the dependent measure (athletic organizational effec- tiveness) and nonintercorrelated. The regression model tested in this study con- sisted of six independent variables (predictors), which were entered sequentially in a predetermined logical order to determine whether they singly or jointly con- tributed to the prediction of athletic organizational effectiveness.

The correlation matrix (see Table 3) provides insight into the goodness-of- fit test as it relates to the predictive power of the set of predictors used in this study. Pearson product-moment correlation coefficients reveal moderate corre- lation (r= .38) between initiating structure as pereeived by the athletic directors (XjIS) and athletic organizational effectiveness (YAOE). The interaction of consideration and initiating structure (XjINT) as perceived by athletic directors is also moderately correlated with the dependent measure (r=.34). However, closer scrutiny reveals that this interaction predictor (X5INT) is also highly cor- related (r=.88) witii one of its individual members (XiCON), indicating mul-

Athletic Director Leader Behavior 169

Table 3

Correlation Matrix

Variables 1 2 3 4 5 6 7

Athletic organizational effectiveness (YAOE) 1.00 —

Consideration as perceived by athletic directors (X,CON) .22 1.00 —

Initiating structure as perceived by athletic directors (XgIS) .38 .50 1.00 —

Consideration as perceived by subordinates (X3CON) .06 .20 .07 1.00 —

Initiatirtg structure as perceived by subordinates p<^IS) - .01 - .06 - .08 .57 1.00 —

Interaction ol con. and IS as perceived by athletic directors (X5INT) .34 .88 .85 .15 - . 1 1 1.00 —

Interaction of con. and IS as perceived by subordinates (XgINT) .04 .11 .00 .92 .83 .05 1.00

ticollinearity. When variables are multicollinear, they are almost linear combinations of each other and are referred to as "near-singularities" (Nonisis, 1983). Other predictors (X,CON, X3CON, XJS, XJNT) had low correlations with the dependent measure (r=.22, .06, - . 0 1 , .04, respectively), indicating that these variables would contribute little to the predictive power of the regres- sion model. Relatively strong intercorrelations between the predictor variables and relatively weak correlations between the predictor and dependent measures are strong indicators of poor predictive p)ower.

Building the regression model required entering the predictors in pre- determined logical steps. A summary of the steps in buildii^ the regression model and the multiple regression analyses are detailed in Table 4. Steps 1 and 2 of the process entered initiating structure and consideration dimensions of leader behavior as perceived by the athletic directors themselves. Steps 3 and 4 entered these same two dimensions as perceived by the selected subordinates. Steps 5 and 6 entered two interaction variables, the product of consideration and initiat- ing structure as perceived by athletic directors and the selected subordinates, to assess their joint contributions to the prediction of athletic organizational effec- tiveness. At each step in building the regression model, the infiuences of previ- ously entered leader behavior dimensions were controlled to assess the "pure" contribution of each subsequently entered variable.

Of the six independent variables cotisidered to predict athletic organiza- tional effectiveness, only initiating structure as perceived by the athletic direc- tors themselves contributed significantly to the preidiction of atiiledc organizational effectiveness. The multiple R (.38) for this variable, which is equivalent to the zero-otider correlation between this predictor and the dependent measure, accounts

170 Branch

Tabie 4

Statistical Suiranary for the Multipie Regression Analyses

Step

1 ? 3 4 5 6

Variable

XgIS X^CON X3CON X^IS ZXjINT ZXgINT

Multiple R

.38

.38

.38

.38 — .38

.14

.14

.14

.14 — .14

R' chEuige

.1403

.0015

.0008

.0002 —

.0001

Fchange

11.260 .116 .060 .012 — .011

Prob. (F)

.0006*

.3675

.4037

.9130 —

.9160

S.E.

.42

.42

.42

.43 — .43

Beta

.3746

.441

.0283

.0156 —

.1160

Correl.

.38

.22

.06 - .01 — .04

*p<.01, one tailed.

for or explains 14% ( ^ =. 14) of the variation in athletic organizational effective- ness. The F test was statistically significant F= .0(X)6, p<.Ol, indicating that this variable was a significant predictor of athletic organizational effectiveness.

In Steps 2 through 6, other predictors were entered into the model to assess their relative contributions. The results indicated that no otber predictor was a significant contributor to the regression model. The multiple R and /?' statistics did not change in these subsequent steps (/?=.38, R^ = .H). The change in /P statistic further reveals how small the pure contributions were beyond the first predictor (.0015, .0008, .0002, .0001). In Step 5, the relative contribution of the interaction predictor was so insignificant that the tolerance level (less than .0001) of the SPSS (X) regression program "disallowed" this variable's entry into the model.

The multiple regression model tested thus did not represent a good fit to the prediction of athletic organizational effectiveness, since only 14% of the depaident measure was predicted by the six independent variables, hi other words, 86% of the athletic directors' perceptions of their program's effectiveness was attributable to something other than their leadership bdiavior as perceived by themselves and/or their subordinates.

Discussion The finding that initiating structure as perceived by the athletic directors them- selves was a significant contributor to the prediction of athletic organizational effectiveness was not surjHising. Chelladurai (1985) r^arts that, within the context of Fiedler's Contingency Model of Leader Effectiveness, a contemporary leader- ship paradigm, the situation in athletic organizations is fiavorable to the extent that an autocratic behavior thf̂ predisposes the leader to task accomplishment is the most effective leadership behavior. Fiedler defines situational fevorable- ness as the "degree to which the situation permits or facilitates the exercise of influence by ttie leader" (Chelladurai, 1985, p. 147).

The fact that consideration was not a significant contributor to the effective-

Athletic Director Leader Behavior 171

ness of the organization was surprising. Contemporary ideas about leadership suggest that both an orientation toward task and goal accomplishment and a development of the human element within the organization contribute to individual and organizational health and effectiveness (Bruno & Lippin, 1979; Mott, 1972; Peters & Austin, 1985; Peters & Waterman, 1982). Many behavioral researchers agree that effective organizations should have leaders v^o balance their efforts and energies between getting things done and developing and nurturing their people (Blake & Mouton, 1964; Fiedler, 1967; Hershey & Blanchard, 1977). Though these dichotomous behaviors may appear mutually exclusive, attention to bofti seems crucial to the effective and productive functioning of an organization.

The finding that subordinate perceptions of leader behavior were not a significant contributor to the prediction of athletic organizational effectiveness was not unexpected. Associate/assistant athletic directors may view their athletic directors with more cynicism, as their own upward mobility is controlled by those athletic directors. In addition, the pressure to win and stay clean in big-time college athletics may preclude the nurturing of personal relationships within the athletic organization. Thus, the incongruency of leadership perceptions may result from a disdain of athletic director control.

The findings in this study represent an exclusive insight into Division I-A athletics. As a result, these findings could not be generalized to a larger athletic population such as other divisions within the NCAA (Division I-AA, n, HI, etc.). Future studies may replicate this study across this and/or other NCAA divisions to compare results. However, each division within the NCAA is distinct in its level of competition, budget, and scholarship restrictions, which makes compari- sons across divisions difficult and generalizations to other divisions statistically invalid.

Earlier leadership researdi in college athletics focused primarily on descrip- tive studies designed to define leadership traits of athletic directors or their leader- ship styles (Austin, 1974; Pruitt, 1976). Though this information does impart some imderstanding of the individual, it does not relate the significance of definhion to the effective functioning of the athletic organization. Little l i^ t has been shed on the rdationship between effective athletic leadership and effective athledc organiza- tional dynamics. Despite their efforts, social scientists are not in agreement on what differentiates an effective leader fixjm an ineffective one in this or other organiza- tional contexts (Hoy & Miskel, 1982; Liken, 1967; Monon, 1983; Mott, 1972).

Some significant issues must be addressed before more definitive conclu- sions can be drawn concerning the relationship between athletic leadership and organizational effectiveness. The measure of athletic effectiveness used in this study must be assessed relative to other effectiveness criteria. The win-at-all-costs objective of some athletic programs may preclude the use of instrum^its designed to measure other aspects of organizational success. Thus, the mission of college athletics—what they want to be or do—should be clarified and standards should be established before effectiveness in this arena can be assessed objectively.

Fur&er inquiry into atiiletic leadership may help to address whether the leadership behavior dimraisions used in this study were f^propriate to measure the phenomena in this arena. OIK could argue that classical £4>proaches to defin- ing leadership, and its relationship to effectiveness, in such a unique environ-

lt may limit their utility.

172 Branch

Conclusion Exploring athletic leader/follower relationships is important in providing under- standing and clarification of the interpersonal dynamics within intercollegiate ath- letic organizations. The problems facing intercollegiate athletic programs indicate a need to examine in still more depth those leader behaviors that could have a positive influence on the relationships between athletic directors and their subor- dinates, as well as those behaviors that influence the entire organizations' health and effectiveness. One area of fiiture inquiry may include examining the influence exerted by significant others in the decision-making process in college athletics today. Such an inquiry may shed some light on who is really in charge (boosters, powerful coaches, college presidents, athletic directors, corporate sponsors, network television, etc.) versus who should be in charge. These and odier questions concerning the future of big-time college athletics are being asked today more frequently than ever. The future of college sports in America may well depend on our success in finding appropriate answers to these questions.

References Aceto, T., (feRyckman, R. (1972). The role of the athletic director. .4«W«;c/i<*ranuff-anCT!,

6, 21. Austin, D. (1974). Leader behavior perceptions and interpersonal needs of atfiledc directors

(Doctoral dissertation. University of Utah, 1973). Dissertation Abstracts Inter- national, 34, 4828A^829A.

Austin, D. (1975). Perceptions and needs of athletic directors. Athletic Administration, 10, 15-16.

Barcikowski, R.S. (1980). Regression analysis. Unpublished manuscript. Blake, R., & Mouton, J. (1964). The managerial grid. Houston: Gulf Publ. Bruno, J., & Lippin, P. (1979). Nice guys finish first (Report No. 79011868). International

Standard Serial No. 0001-8376. Buros, O.K. (1978). The eighth mental measurements yearbook. Highland Park, NJ: The

Gryphon Press Chelladurai, P. (1985). Sport management. London, Ontario: Spons D>'namics. Dulac, G. (1989, February). Athletic directors merge games with business, lhe Pittsburgh

Press, pp. Bl, B4. Fiedler, F.E. (1967). A theory of leadership effectiveness. New York: McGraw-Hill. Halpin, A.W. (1957). Manual forthe leader behavior description questionnaire. Colum-

bus: The Ohio State University, Bureau of Business Research, College of Com- merce and Administration.

Hatfield, B., Wrenn, J., & Bretting, M. (1987). Comparison of job responsibilities of intercollegiate athletic directors and professional general managers. Journal of Sport Management, 1, 129-145.

Hershey, P., &Blanchard, K. (1977). Management of organizational behavior: Utilizing human resources (3rd ed.). Englewood Cliffs, NJ: Prentice-Hall.

Hoy, W.K., & Miskel, C.G. (1982). Educational administration: Theory, research and practice (2nd ed.). New York: Random House.

Krupa, G., & Dunnavant, K. (1989, January 2). The struggle with the downside. SPORTS inc., pp. 33-38.

Athletic Director Leader Behavior 173

Landry, D, (1983), What makes a top college athletic director. Athletic Administration, 18, 20.

Likert, R. (1967), The human organiTjotion: Its management and value. New York; McGraw-Hill,

McGee, M, (1984), Management approach: Training athletic directors. Athletic Adminis- tration, 18, 20,

Morton, M.K, (1983), A study of presidential leadership behavior and perceived organiza- tional effectiveness at selected four-year public historically black colleges and univer- sities (Doctoral dissertation, Kent State University). Dissertation .Abstracts Inter- national, 44, 2993A.

Mott, P.E. (1972). The characteristics of effective organizations. New York: Harper & Row, Nonisis, M.J. (1983). SPSSPC) introductory statistics guide. New York: McGraw-Hill. Peters, T,, & Austin, N, (1985), A passion for excellence. New York: Random House. Peters, T., & Waterman, R, (1982), In search of excellence. New York: Harper & Row, Pruitt, P.J. (1976). Perceived leader behavior of male and female intercollegiate athletic

directors in the United States (Doctoral dissertation. University of Illinois at Urhana- Champaign). Dissertation Abstraas Intemationai, 60, 6199A,

Stogdill, R.M, (1948). Personal factors associated with leadership: A survey of the litera- ture. Journal of Psychology, 25, 35-71,

Stogdill, R,M, (1963). Manual for the leader behavior description questionnaire—Form XII: An experimental revision. Columbus: The Ohio State University, Bureau of Business Research, College of Commerce and Administration.

Stogdill, R.M. (1969), Validity of leader behavior descriptions. Personnel Psychology, 22, 153-158.

Stogdill, R.M., Goode, O.S., & Day, D.R. (1962). New leader behavior description sub- scale. Journal of Psychology, 54, 259-269,

Sziiagyi, A.D., & Keller, R.T, (1976), A comparative investigation of the supervisory behavior description questionnaire (SBDQ) and the revised leader behavior descrip- tion questionnaire (LBDQ-Form XII). Academy of Management Journal, 19, 642-649,

25 Institutions under NCAA sanctions, (1989, February 15). The Chronicle of Higher Education, p. A36,