An analysis of donor involvement, gender, and giving in college athletics.

Shapiro, Stephen L. ; Ridinger, Lynn L.

Introduction

Charitable contributions continue to represent a significant portion of revenue for college athletic departments. According to Fulks (2008), athletic donations account for 31% of generated revenue at Football Bowl Subdivision (FBS) institutions. These findings were similar for Football Championship Subdivision (FCS) institutions, where 29% of generated revenue came from individual donations. In 2006, the median annual contribution was $5,826,000 and $635,000 for FBS and FCS institutions, respectively. Donations, along with ticket sales, have become a primary source of generated revenue (not including revenue allocated to the athletic department by the university) (Fulks, 2008).

Furthermore, the current economic climate has created a challenging environment for all nonprofit organizations. In the United States, charitable contributions decreased by 5.7% in 2008 (Giving USA, 2009). In tough economic times, the need for fundraising dollars continues to increase in college athletics. Thus, understanding donor behavior has become vitally important. There has been a steady increase in research on college athletic donor behavior over the past few decades. The majority of investigations have focused on the influence of winning on charitable contributions (Coughlin & Erekson, 1985; Daughtrey & Stotlar, 2000; Humphreys & Mondello, 2007; Sigelman & Carter, 1979; Sigelman & Bookheimer, 1983), donor motivations (Billing, Holt, & Smith, 1985; Gladden, Mahony, & Apostolopoulou, 2005; Mahony, Gladden, & Funk, 2003; Staurowsky, Parkhouse, & Sachs, 1996), and the relationship between athletic and academic giving (Stinson & Howard, 2004, 2007). However, research on the concept of involvement and athletic donor behavior is limited (Tsiotsou, 2004).

Additionally, although athletic departments have been generally successful generating revenue through fundraising, it appears that the ability to cultivate female donors is lacking (Tsiotsou, 2006). Women represent a viable portion of the donor population that has the potential to generate significant revenue through voluntary support. Over the past few decades, there has been enormous growth in the economic potential of women as evidenced by the fact that an increasing number of women are participating in the American workforce, starting businesses, and earning advanced degrees (Shaw, 1992). Women control a large percentage of the nation's wealth and constitute an untapped market for fundraisers (Verner, 1996). Women live an average of seven years longer than men so they are often in charge of estates after their spouse dies, a fact that may be of particular importance to nonprofit organizations (Braus, 1994). Despite all of the reasons to indicate a great increase in female donor potential, scant attention has been given to women's philanthropy in the sport management literature (Tsioutsou, 2006).

Empirical research in the nonprofit sector has found gender differences in donor behavior. In comparison to men, women are more likely to volunteer at charitable organizations before giving and they desire closer relationships with the charities they support (Kaplan & Hayes, 1993; Shaw, 1992; Sommerfeld, 2000). Men tend to give to charities to enhance their own standing or to gain access to social or tangible rewards such as invitations to special events, while women give more to "people" charities to promote social change or help others less fortunate (Kottasz, 2004; Newman, 2000). Women are more likely than men to ask questions and acquire information before they are comfortable about making substantial gifts (Newman, 2000). Although Mesch, Rooney, Steinberg, and Denton (2006) found that single women are more likely to be donors than single men, and Andreoni and Vesterlund (2001) discovered that women appeared more altruistic than men when the price of giving is high, most of the evidence suggests that women give less than men when making charitable contributions (Hall, 2004). "Some researchers have argued that small gifts by women who can easily afford to give more are rooted in insensitive fundraising practices that ignore women's contributions, reflect male rather than female priorities, and exclude women from top leadership positions" (Hall, 2004, p. 73). A better understanding of gender differences in regard to giving behavior can be used by fundraisers to develop strategies to better meet the needs of female donors.

Changes in the economic profile of women over the past few decades along with growth in female participation and interest in college sport (Zgnoc, 2010) make women donors an attractive market segment for intercollegiate athletic fundraisers. On college campuses, women account for more than half of undergraduates today (Strout, 2007). These undergraduates will become alumni, who represent the largest donor base for academic institutions. However, the vast majority of athletic donors appear to be male. Recent studies (Mahony et al., 2003; Tsiotsou, 2006) have shown that female donors account for 25% or less of individual athletic donors at various institutions. Many researchers agree that priority seating for athletic events is the key motive for making contributions to intercollegiate athletics (Gladden et al., 2005; Mahony et al., 2003; Stinson & Howard, 2004); however, these studies did not examine gender differences in motives. The few studies that have focused on gender and athletic fundraising have found that women are less motivated by the social and tangible benefits associated with athletic gifts (Staurowsky, 1996; Tsiotsou, 2006) and are more motivated by philanthropic concerns (Staurowsky, 1996; Verner, 1996). These findings, coupled with research on gender differences from the nonprofit sector, suggest that athletic fundraisers may need to modify their approaches to cultivate greater involvement of female donors.

The concepts of involvement and donor gender are two areas that are underdeveloped in the college athletic fundraising literature. Therefore, the primary purpose of this study was to examine involvement and its relationship to donor gender. The involvement construct was examined to identify any gender differences that exist. Additionally, gender differences in donations (annual contributions and donor length) and pertinent demographics (age and income) were assessed. Five research questions were developed to guide the current study:

Research Question #1: Does involvement differ between male and female college athletic donors?

Research Question #2: Do male and female donors differ in annual contribution amount?

Research Question #3: Do male and female donors differ in donor longevity?

Research Question #4: Do male and female donors differ in age?

Research Question #5: Do male and female donors differ in annual income?

Review of Literature

The Involvement Construct

The concept of involvement was first introduced in psychology as part of social-judgment theory (Sherif & Cantril, 1947; Sherif & Hovland, 1961). Involvement has generally been defined in social-psychological terms as an unobservable state of motivation, arousal, or interest between an individual and an activity or product (Rothchild, 1984). Involvement, however, extends beyond individual motives and mere participation; it looks at the relevance or meaning of an activity or product within the context of an individual's overall outlook on life (Wiley, Shaw & Havitz, 2000). It is seen as an attitude that is relatively enduring in nature and is important to the individual on an ongoing basis. Interest in involvement gained momentum in the consumer behavior field in the 1980s as researchers utilized the concept to understand purchase behavior related to consumer products (Laurent & Kapferer, 1985; Rothchild, 1984; Zaichkowsky, 1985).

The majority of involvement research in consumer behavior has focused on the level of involvement (low involvement vs. high involvement) of consumers and its effect on decision making, information gathering, and information sources. According to Zaichkowsky (1985), involvement focuses on personal relevance. There are three major factors that affect a person's involvement level: 1) characteristics of the person, 2) characteristics of the product, and 3) characteristics of the situation. These characteristics ultimately influence consumer behavior and purchase intentions, and serve as the basis for Zaichkowsky's (1985) original 20-item Personal Involvement Inventory (PII). Due to the number and redundancy of items, Zaichkowsky (1994) simplified and refined the PII by reducing the scale to 10 total items with two dimensions (cognitive and affective). Cognitive involvement stresses a person's information processing, whereas affective involvement is focused on a person's feelings. The PII is a semantic-differential scale using adjectives to describe involvement concepts. The items captured in the affective dimension were: Interesting, Exciting, Appealing, Fascinating, and Involving. The items captured in the cognitive dimension were: Needed, Important, Relevant, Means A Lot, and Valuable.

The PII has been used as a measure of consumer involvement for products, advertisements, and purchases, but there has been limited investigation in relation to services. Stafford and Day (1995) extended Zaichkowsky's (1994) work through an investigation of involvement within the context of service research. The authors suggested that both cognitive and affective components of consumer involvement exist in services, and the 10-item PII was an appropriate measure of involvement within a service context. Celuch and Taylor (1999) also investigated the efficacy of Zaichkowsky's (1994) PII inventory within the context of service research. The authors reexamined the PII scale across multiple service organizations in an effort to provide support across a variety of industries. The results of this study provided strong support for a further reduced 8-item version of the PII. Celuch and Taylor dropped the Interesting and Involving items from the Zaichkowsky 10-item PII. Based on the sample scores, the modified 8-item PII captured both cognitive and affective factors identified in previous research (Zaichkowsky, 1994). Reliability (coefficient alpha) scores were satisfactory, ranging from .82 to .86 for affective involvement and .80 to .93 for cognitive involvement, respectively.

Donor Involvement and College Athletic Fundraising

Tsiotsou (1998) developed the Giving to Athletics Model (GAM) in an effort to explain why individuals make contributions to athletic programs. Of the seven proposed independent variables in the GAM, only involvement and emotional motivation were significant in directly explaining donations to athletics. The author concluded that involvement should be used in future attempts to understand donor behavior. Tsiotsou (2004) extended this research by attempting to classify the giving level of donors based upon income and level of involvement (high or low). The 10-item version of Zaichkowsky's (1994) PII scale was used in this study to measure involvement with athletics. The findings showed that involvement was a discriminating factor in determining donation amount. High-income, high-involvement donors were more likely to make large contributions to athletics.

Most recently, Tsiotsou (2006) focused on college athletic donor gender and involvement. The construct of involvement (along with income, donor motives, annual contributions, spectator attendance, and sport experience) was examined to identify differences between male and female donors. The Zaichkowsky (1994) 10-item PII scale was also used in this investigation to measure involvement. The results showed no significant difference in level of involvement between male and female donors. It should be noted, however, that involvement was treated as a unidimensional scale in Tsiotsou's (2004, 2006) investigations. There was no attempt to examine the cognitive and affective involvement facets of the PII separately. Also, there was no assessment of the scale's properties to ensure the appropriateness of the PII for college athletic donors.

Based on the aforementioned studies on donor involvement in college athletics, it appears that involvement may have an impact on donation amount, but there is no evidence of any gender differences. However, measurement and scale selection issues existed in the assessment of the PII in these examinations. First, the involvement construct has been identified as multidimensional, yet Tsiotsou's (2004, 2006) results were based on a unidimensional interpretation of the PII. The cognitive and affective factors offer unique facets of involvement, which may provide different results when measured separately. Second, reliability and validity were not examined prior to analysis of the PII to provide evidence that it is an appropriate measure of athletic donor involvement. Lastly, the previous studies on donor involvement used the Zaichkowsky (1994) 10-item PII as opposed to the condensed 8-item version of the PII, modified by Celuch and Taylor (1999). Since the 8-item version of the PII represents a more parsimonious measure of service involvement, it may be the most appropriate assessment of donor involvement.

Gender and Giving

Tsiotsou (2006) is one of a very limited number of studies to focus on gender differences in those who donate to intercollegiate athletic programs. Her findings revealed that income, donation amount, specific donor motives, sport experience, and attendance were variables that contributed significantly to the discrimination between female and male athletic donors, but involvement was not a significant factor. In an earlier study on women and athletic fundraising, Staurowsky (1996) found that female athletic donors appear to be younger than male donors, contribute less money, and are more inclined to give to women's athletic programs. Furthermore, women donors were not as motivated as men by the material gain associated with the act of giving or by the social interaction and approval related to being part of an athletic support group.

Verner (1996) reviewed the literature pertaining to women's philanthropy and provided ideas for intercollegiate athletic programs to cultivate female donors. Keying in on studies by Shaw and Taylor (1995) and Stone and Sublett (1992), both based on qualitative data collected from interviews and focus groups with women philanthropists, Verner outlined several reasons why women make charitable contributions. Recurring themes associated with women's giving that emerged included personal commitment, volunteer involvement, and strong feelings about a cause or charitable organization. Family tradition also was an influencing factor as most of the participants in these studies had family role models, particularly mothers, who donated to charities. In addition, women expressed a sense of responsibility or desire to "give back" to meaningful causes as well as a need to bring about change and make a difference. Verner's findings were based primarily on information from private donor and philanthropic activity within the nonprofit and political sectors due to the dearth of literature on women as financial donors to intercollegiate athletics.

Hall's (2004) investigation of gender differences in giving also focused on philanthropy in the nonprofit sector. Hall considered three observations about sex differences in giving: 1) women's gifts tend to be smaller than men's gifts, 2) it takes longer to cultivate significant gifts from women, and 3) unlike men, women do not give competitively or to receive perks; however, she noted that there are few large scale empirical studies to support these claims. Women seem to undervalue their giving ability and make fewer headline-grabbing gifts in comparison to their male counterparts (Hall, 2004). Women are more likely than men to volunteer before giving and seek closer contact with the charities they support (Kaplan & Hayes, 1993; Shaw, 1992; Sommerfeld, 2000). Also, it may take longer to cultivate significant gifts from women. Women tend to ask more questions than men and take more time in deciding to make a sizable gift. Some fundraisers attribute this hesitation in giving to women's lack of financial skills or fear they will outlive their money (Hall, 2004). Women, more so than men, want to know how their charitable dollars are being used, and view charity as a means to secure additional friendships and involvement in the community (Marx, 2000). Those in the fundraising profession have asserted that women tend to give to promote social change or help others less fortunate whereas men give for the recognition and status (Newman, 2000). It has been suggested that unlike men, women are not motivated by competition with their peers to make the largest gift, nor are they interested in having buildings named after them (Taylor & Shaw, 1997). Hall (2004), however, noted that examples of competitive female donors and women who seek out perks for giving are on the rise. Thus, there is still much to be learned about gender-based differences in charitable giving.

The donor characteristics and gender differences identified by the previous literature present opportunities for further investigation. First, some of the female donor characteristics presented in the literature suggest motivations based on personal preferences. According to Zaichkowsky (1985), personal preferences are defined as "inherent interests, values, or needs that motivate one toward the object" (p. 342). Personal preferences are a primary component of involvement in terms of purchase intentions. Therefore, level of involvement may be a key indicator in understanding donor behavior for males and females. Second, results from previous investigations have shown that females contribute less money compared to their male counterparts. The difference in the gift amounts could be a result of a female donor population who are younger and earn less income than male donors (Staurowsky, 1996); however, additional investigation is warranted to provide a more current understanding of donor dynamics and trends related to gender. Thus, this current study was developed to investigate differences between male and female college athletic donors in terms of involvement, annual contribution amount, donor longevity, age, and annual income.

Methods

Sample

The population for the current study consisted of current college athletic donors. An online survey was sent to 7,467 current donors from three NCAA Division I FBS institutions located in the mountain and southwest regions of the United States. All three universities compete in the same athletic conference. Two of the institutions are public and one is private, and university enrollment ranges from 9,000 to 28,000 students. Three institutions were chosen in order to collect a large enough sample of current female donors for data analysis. A total of 1,664 usable surveys were returned for a response rate of 22.2%. The majority of respondents (N = 1,664) were male (77.3%), which was consistent with the donor gender breakdown at each of the three institutions being examined and previous gender examinations in college athletics (Tsiotsou, 2006). Table 1 provides a sample breakdown by institution.

Instrumentation

The questionnaire used for the current study consisted of three sections with a total of 19 items. The first section focused on donor information items such as donation amount, donor level, and total years as a donor. The second section was comprised of an adapted version of the Celuch and Taylor (1999) reduced, 8-item PII originally developed by Zaichkowsky (1985, 1994). Examples of the semantic scale item anchors include "important--unimportant" and "exciting unexciting." Scale items were measured from 1 = low involvement to 5 = high involvement. The adapted PII had two subdimensions of involvement: cognitive involvement, which is made up of five items (Needed, Important, Relevant, Means A Lot, and Valuable) and affective involvement, which is made up of three items (Exciting, Appealing, Fascinating). The 8-item modified PII has shown good reliability in previous examinations in service related industries with alpha values ranging from .80 to .92 (Celuch & Taylor, 1999). The final section of the survey focused on demographic items in order to profile the typical donor at the institutions being examined.

Procedure

Questionnaires were administered through an online format. Each institution's athletic department sent an email blast out to all current donors. Each potential participant received an introductory email explaining the purpose of the study along with a link to the web-based survey. A follow up email was sent to all potential participants two weeks later in an effort to increase response rate. In addition, each of the three athletic departments gave respondents the option of entering a drawing to win passes to an upcoming athletic department event. This information was kept separate from survey responses to maximize anonymity and confidentiality.

Data Analysis

A confirmatory factor analysis (CFA) was initially conducted on the PII to examine the factor structure of the involvement construct based on the pooled sample of current donors. Previous theory on involvement and scale development of the PII (Celuch & Taylor, 1999; Zaichkowsky 1985, 1994) drove specification of the factor model. Therefore, CFA was the most appropriate factor analytic technique (Brown, 2006).

Multiple measures of fit were used to examine the factor structure of the PII. Overall goodness of fit was assessed using a robust chi-squared test; however, according to Hu and Bentler (1999), this test can be sensitive to sample size and should not be used exclusively in determining model fit. Therefore, standardized root mean square residual (SRMR), root mean square error of approximation (RMSEA), and the comparative fit index (CFI) were examined to provide additional sources of fit that are widely accepted in applied research and have shown satisfactory performance in model simulation analyses. According to Hu and Bentler, SRMR values close to .08 or below, RMSEA values close to .06 or below, and CFI values close to .95 or greater provide evidence of an adequate model fit. Additionally, average variance extracted (AVE) was assessed for validity-related evidence and alpha coefficients were examined within each factor of the PII in order to assess reliability-related evidence. Means and standard deviations were subsequently calculated for each subdimension of involvement.

Multiple statistical procedures were conducted to answer the five proposed research questions. In order to examine potential gender differences in cognitive and affective donor involvement (RQ1), a one-way multivariate analysis of variance (MANOVA) was conducted. Assumptions of normality, homogeneity of variance/covariance matrices, and independence were considered when conducting the MANOVA test. In addition, MANOVA assumes that there is a linear relationship (linearity) between the dependent variables in the model (Tabachnick & Fidell, 2007). The data presented no apparent violations of MANOVA assumptions.

A descriptive discriminate analysis (DDA) was used as a post-hoc procedure to identify the dependent variable that maximally discriminates among the groups associated with the independent variable (Duarte Silva & Stam, 1995). In the current study, DDA was used to examine which of the involvement constructs was the most important discriminator of gender. An analysis of the structure matrix in DDA provided specific information regarding which dependent variable correlated highest with the linear combination of dependent variables and, therefore, is a more important discriminator among male and female donors (Tabachnick & Fidell, 2007).

Independent t-test procedures were conducted to examine potential gender differences in annual contribution, donor longevity, and donor age (RQs 2-4). Assumptions of normality, independence, and homogeneity of variance were considered prior to the independent samples t-tests. Levene's Test for Equality of Variances was significant for each set of t-tests; therefore, a Welch's t-test was conducted for each of these research questions.

Finally, a chi-square analysis was conducted to identify potential differences in income level between male and female donors (RQ5). Assumptions of independence, exhaustiveness and mutual exclusivity, and minimum cell size were considered prior to chi-square analysis. No assumption violations were found. A significance level of .05 was set for the MANOVA and chi-square procedures. However, due to the use of three independent samples t-tests, a Bonferroni adjustment was used to control for Type I error. The adjusted alpha value was set at .017 for these statistical procedures.

Results

Demographic Profile

The average age for male donors (n = 1,286) was 53.5. The majority of male donors were Caucasian (88.6%), married (85.8%), and had an annual household income above $100,000 (62.3%). In addition, 37% of male donor respondents had a graduate degree. The average annual donation for males was $1,360.57 and the average length of annual giving was 11.4 years. The average age for female donors (n = 360) was 53.9. The majority of female donors were also Caucasian (83.9%) and married (65.8%). Furthermore, 44.2% of female respondents had an annual household income above $100,000 and 30.6% had a graduate degree. Finally, the average annual donation for females was somewhat lower than males at $728.76 and the average length of annual giving was 9.5 years. Demographic information was also broken down by institution. Table 1 provides a breakdown of donor characteristics segmented by institution.

Confirmatory Factor Analysis

CFA was conducted on the modified, two-factor, 8-item PII model. The results indicated that the data fit the model well. Absolute fit, parsimony correction, and comparative indices all represented a reasonably good fit: [X.sup.2] (19) = 49.41; p = <.001; RMSEA = .037; SRMR = .025; CFI = 1.0. The final model consisted of two subdimensions of involvement. All t-values were greater than 2.0, which is considered satisfactory (Thompson, 2004). A summary of the anchors, factor loadings, t-values, and standard errors in the final PII structure are presented in Table 2.

Validity and Reliability

Convergent validity was assessed on the PII with reference to AVE. According to Fornell and Larcker (1981), AVE scores above .50 indicate an adequate ratio of total variance that is due to the latent variable. AVE values were .818 and .844 for affective and cognitive factors, respectively. This information provided evidence of the scale's convergent validity. In addition, internal consistency of the cognitive and affective involvement factors was examined with Cronbach's alpha estimates. Internal consistency was above the standard .70 cutoff (Cronbach, 1951) with coefficient alphas of .87 and .92 for affective and cognitive factors, respectively.

Research Question 1

The first research question addressed potential gender differences regarding both subdimensions of involvement. Means and standard deviations were calculated for the affective and cognitive involvement factors. In terms of the overall sample of current donors, cognitive involvement (M = 4.54, SD = .587) had greater scores than affective involvement (M = 4.16, SD = .586). However, a one-way MANOVA was conducted to examine whether these involvement subdimensions significantly differed between male and female donors. Wilk's Lambda approximation to F was reported. MANOVA results indicated a significant gender difference for at least one of the involvement factors F(2,1643) = 3.52, p = .030. A post-hoc DDA was examined to identify which involvement factor significantly differed between male and female donors. A structure (loading) matrix of correlations between predictors and discriminant functions suggested that affective involvement is the best predictor for distinguishing between male and female donors (.85). Based on an assessment of the structure matrix and the standardized discriminant function coefficients, the cognitive involvement variable did not effectively distinguish between male and female donors. A follow up one-way ANOVA was found to be significant F(1,1644) = 5.05, p = .024. Affective involvement means indicated that female donors had a stronger sense of affective involvement (M = 4.24, SD = .715) compared to male donors (M = 4.14, SD = .707).

Research Questions 2-4

The second, third, and fourth research questions addressed potential gender differences regarding annual contributions, donor longevity, and donor age. Table 3 summarizes the mean gender differences for these variables. Three independent samples f-tests were conducted to examine these research questions. Results of the first f-test (RQ2) indicated a significant difference in annual contributions between male and female donors t(1,210.39) = 5.18, p < .001. Male donors contributed approximately 1.9 times more than females on an annual basis. Results of the second f-test (RQ3) indicated a significant difference in donor longevity between male and female donors t(591.49) = 3.46, p = .001. Male donors have been contributors for approximately two more years compared to their female counterparts. Results of the third f-test (RQ4), which examined gender differences in age were not found to be significant.

Research Question 5

The fifth research question addressed potential gender differences in annual income. Table 4 summarizes the mean gender differences for each category of annual income. Results of a chi-square analysis indicated a significant difference between the annual income of male and female donors [X.sup.2] (4) = 55.07, p < .001. The lower income level categories included a larger percentage of female donors compared to male donors. However, in the largest income category ($100,000 and above), which included the vast majority of donors, the percentage of male donors was considerably higher.

Discussion

The purpose of this study was to examine differences between male and female college athletic donors in terms of involvement and several demographic characteristics. Prior to testing for gender differences in involvement, the Celuch and Taylor (1999) modified 8-item PII was assessed to examine the involvement construct among college athletic donors. Due to the fact that the PII has not been appropriately assessed using a sample of college athletic donors, it was important to examine the factor structure as well as reliability and validity-related evidence of this instrument. The 8-item, PII model showed adequate reliability and validity based on the sample scores. Cognitive and affective subdimensions of involvement were clearly identified. This study investigated whether donor involvement and selected demographic variables differed between male and female contributors. The results suggested that significant differences existed for gender. Specifically, affective involvement was stronger for female donors as opposed to male donors. Additionally, in comparison to their male counterparts, female donors made smaller annual contributions, had less donor longevity, and had lower annual income levels. There were no significant differences in the age of donors based on gender.

Theoretical Implications

The reduced, 8-item, PII instrument (Celuch & Taylor, 1999) appears to be applicable in terms of donor involvement based on the current sample scores. These findings demonstrate strong support for use of a parsimonious measure of the construct of involvement. In addition, the results provide support for the use of the PII as a measure of donor involvement within college athletics. The reduced scale appears to capture both cognitive and affective dimensions of involvement within a nonprofit setting. However, while the items performed well within the context of athletic donors in this study, only three schools with similar characteristics were examined. Additional research across a variety of different institutions is suggested to further enhance the generalizability of the PII for college athletic donors.

The current results also provided evidence of gender differences in terms of donor involvement. This discovery was contrary to the findings of Tsiotsou (2006). However, the current study examined both cognitive and affective dimensions of donor involvement as opposed to one unidimensional measure. Only affective involvement had a significant difference between male and female donors. From a theoretical standpoint, this information extends the knowledge base regarding the PII and the construct of involvement. First, the results provide clear evidence of two distinct involvement dimensions, which is consistent with previous involvement research (Celuch & Taylor, 1999; Zaichkowsky, 1994). Second, the significant gender difference within the affective dimension represents a distinction between male and female donors at an emotional level. Affective involvement is focused on emotional and self-image issues that influence attitude formation (Park & Young, 1983). Females appear to have stronger affective involvement which may influence donor attitudes and behavior in a unique fashion. Females may have a stronger sense of involvement through personal relevance based on emotional or aesthetic appeals. Park and Young described these appeals as value-expressive motives. Therefore, female donors may feel more involved through a message that is value-expressive as opposed to utilitarian in nature. These findings are supported by previous research on gender and fundraising, where females tend to contribute based on emotional motivations and/or cues (Kottasz, 2004; Newman, 2000). Third, this study has extended the research on gender and giving in college athletics by providing additional information on several demographic variables. Similar to the results reported in both Staurowksy (1996) and Tsiotsou (2006), this study found that female donors, on average, contribute less than male donors. The annual income level of female donors was less than male donors, supporting the findings from Tsiotsou. Interestingly, the age of donors did not differ significantly based on gender. This result differs from Staurowsky's finding that women donors were younger than male donors. Finally, this current study adds a new element to the literature by finding gender differences in donor longevity, a variable not reported in previous studies

Practical Implications

According to Zaichkowsky (1985), involvement is focused on a consumer's personal relevance to a product. It appears that females have stronger relevance to the cause in terms of emotions and self-image. These findings present an opportunity for athletic departments to develop unique donor marketing strategies focused specifically on the interests of potential female contributors. Prior to making financial contributions, women oftentimes want to be involved with the organizations that they support (Hall, 2004). It may be prudent for athletic departments to develop opportunities for involvement for women prior to soliciting donations. Meet and greet interactions with coaches and players provide a way for female fans to get to know the school's athletic teams and feel more emotionally attached to them. Also, luncheons or special events for potential and current donors would contribute to affective involvement by enabling female donors to develop relationships with other supporters of the athletic program.

Another tactic to build the base of female donors is to target women in middle income brackets. In this study, the percentage of women donors in the middle income brackets (i.e., $30,000-49,999, $50,000-69,999, and $70,000-99,999) was higher than the percentage of men. The majority of male donors (65.4%) earned more than $100,000 annually; however, less than half of female donors (47.6%) were in the highest income bracket. Although the gift amount from women in middle income brackets may be less than wealthier individuals, the additive effects of greater numbers can be substantial. It is also important for athletic fundraisers to set goals for increasing the longevity of female donors. They should strive to get younger women involved in annual giving and work hard to meet their needs so that the donor relationship is sustained and can accrue over time.

Although there were no gender differences in cognitive involvement, it is important to note that overall scores for cognitive involvement (M = 4.54, SD = .587) were greater than scores for affective involvement (M = 4.16, SD = .586). In a quest to increase female donors, athletic fundraising personnel should not focus exclusively on affective involvement. Increasing both types of involvement will likely result in higher levels of giving as there is a correlation between involvement and donation amount (Tsiotsou 1998, 2004). That is, high involvement donors are more likely to make larger contributions to athletics. Increases in cognitive involvement can be garnered by ensuring donors that their support of the athletic program is needed, important, relevant, meaningful, and valuable. This can be done through effective communication so that donors understand how their gifts contribute to athletics and the development of student-athletes. This is especially important to women who, more so than men, want to know how their charitable dollars are being used (Marx, 2000). Based on the findings of Staurowsky (1996) and Tsiotsou (2006), marketing efforts targeting female athletic donors may be more effective if they highlight intangible philanthropic benefits such as the opportunity to help student-athletes rather than focusing on tangible benefits and perks such as priority seating and preferred parking. Being sensitive to the involvement needs of female donors may allow athletic fundraisers to better leverage this potentially lucrative market segment.

Future Research

There are multiple opportunities for future investigations of the influence of involvement on charitable contributions in college athletics. As mentioned previously, additional assessment of the PII using athletic donors from diverse institutions (i.e., size of school and athletic department, geographic location, level of competition) will enhance generalizability for the population of college athletic donors. Future research should also focus on the influence of both cognitive and affective involvement on general donor behavior (i.e. decisions to contribute, gift amount, retention, and longevity). Understanding the impact of involvement on past donor behavior and future intentions will provide additional evidence of the importance if this construct. Additionally, there may be some interaction between donor motivations and donor involvement. These two attitudinal measures could be examined within the same sample of donors to assess the relationship between the two constructs.

Another idea for future research would be to examine perceptions of importance of the various benefits and required giving levels offered by intercollegiate athletic departments and compare these by gender. Although some studies have found that women are less motivated by the social and tangible benefits associated with contributions to athletic programs (Staurowsky, 1996; Tsiotsou, 2006), there is some evidence that more women are seeking out perks and recognition for giving (Hall, 2004). There is still much to be learned about gender differences in donor behavior. The potential population of female donors will continue to grow; therefore, an increased understanding of male and female donor attitudes will ultimately enhance recruitment and retention strategies for development offices.

References

Andreoni, J., & Vesterlund, L. (2001). Which is the fair sex? Gender differences in altruism. Quarterly Journal of Economics, 116(1), 293-312.

Billing, J. E., Holt, D., & Smith, J. (1985). Athletic fund-raising: Exploring the motives behind donations. Chapel Hill, NC: University of North Carolina Press.

Braus, P. (1994). Will boomers give generously? American Demographics, 48, 50, 52, 57.

Brown, T. A. (2006). Confirmatory factor analysis for applied research. New York, NY: The Guilford Press.

Celuch, K., & Taylor, S. A. (1999). Involvement with services: An empirical replication and extension of Zaichkowsky's personal involvement inventory. Journal of Consumer Satisfaction, Dissatisfaction, and Complaining Behavior, 12, 109-122.

Coughlin, C.C., & Erekson, O. H. (1985). Contributions to intercollegiate athletic programs: Further evidence. Social Science Quarterly, 66(1), 194-202.

Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16(3), 297-334.

Daughtrey, C., & Stotlar, D. K. (2000). Donations: Are they affected by a football championship? Sport Marketing Quarterly, 9(4), 185-193.

Duarte Silva, P. A., & Stam, A. (1995). Discriminant analysis. In L. G. Grimm & P. R. Yarnold (Eds.), Reading and understanding multivariate statistics (pp. 277-318). Washington, DC: American Psychological Association.

Fornell, C., & Larcker, D. (1981). Evaluating structural equation models with unobservable and measurement error. Journal of Marketing Research, 18, 39-50.

Fulks, D. L. (2008). 2004-2006 NCAA revenue and expenses of Division I intercollegiate athletic programs report. Indianapolis, IN: National Collegiate Athletic Association.

Giving USA (2009, June). U.S. charitable giving estimated to be $307.65 billion in 2008. Glenview, IL: Author.

Gladden, J. M., Mahony, D. F., & Apostolopoulou, A. (2005). Toward a better understanding of college athletic donors: What are the primary motives? Sport Marketing Quarterly, 14(1), 18-30.

Hall, H. (2004). Gender differences in giving: Going, going, gone? New Directions for Philanthropic Fundraising. 43, 71-81.

Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1-55.

Humphreys, B. R. & Mondello, M. (2007). Intercollegiate athletic success and donations at NCAA Division I institutions. Journal of Sport Management, 21(2), 265-280.

Kaplan, A. E., & Hayes, M. J. (1993). What we know about women as donors. In A. J. von Schlegell & J. M. Fisher (Eds.), New directions for philanthropic fundraising Women as donors, women as philanthropists (pp. 5-20). San Francisco: Jossey-Bass.

Kottasz, R. (2004). Differences in the donor behavior characteristics of young affluent males and females: Empirical evidence from Britain. International Journal of Voluntary and Nonprofit Organizations, 15(2), 181-203

Laurent, G., & Kapferer, J. N. (1985). Measuring consumer involvement profiles. Journal of Marketing Research, 22, 41-53.

Mahony, D. F., Gladden, J. M., & Funk, D. C. (2003). Examining athletic donors at NCAA Division I institutions. International Sports Journal, 7(1), 9-27.

Marx, J. D. (2000). Women and human services. Social Work, 45, 27-38.

Mesch, D. J., Rooney, P. M., Steinberg, K. S., Denton, B. (2006). The effects of race, gender, and marital status on giving and volunteering in Indiana. Nonprofit and Voluntary Sector Quarterly, 35(4), 565-587.

Newman. R. (2000). Gender differences in philanthropy. Fund Raising Management, 28-29.

Park, C. W., & Young, S. M. (1983). Types and levels of involvement and brand attitude formation. In R. P Bagozzi & A. M. Tybout (Eds.), Advances in consumer research, 10, 320-324, Ann Arbor, MI: Association for Consumer Research.

Rothchild, M. L. (1984). Perspectives on involvement: Current problems and future directions. Advances in Consumer Research, 11, 216-217.

Shaw, S. C. (1992). Communicating with women: Understanding and applying the differences. In A. J. von Schlegell & J. M. Fisher (Eds.), New directions for philanthropic fundraising: Women as donors, women as philanthropists. San Francisco: Jossey-Bass.

Shaw, S. C., & Taylor, M. A. (1995). Reinventing fundraising Realizing the potential of women's philanthropy. San Francisco: Jossey-Bass.

Sherif, M., & Cantril, H. (1947). The psychology of ego involvements, social attitudes and identifications. New York: Wiley.

Sherif. M., & Hovland, C. I. (1961). Social judgment: Assimilation and contrast effects in reaction to communication and attitude change. New Haven, CT: Yale University Press.

Sigelman, L., & Bookheimer, S. (1983). Is it whether you win or lose? Monetary contributions to big-time college athletic programs. Social Science Quarterly, 64(2), 347-359.

Sigelman, L., & Carter, R. (1979). Win one for the giver? Alumni giving and big-time college sports. Social Science Quarterly, 60(2), 284-294.

Sommerfeld, M. (2000, October 5). A close look at ways to increase giving by women and to instill the habit in girls. Chronicle of Philanthropy, p. 12.

Stafford, M. R., & Day, E. (1995). Measuring service involvement: A preliminary assessment. American Marketing Association Educators' Proceedings, 6, 75-80.

Staurowsky, E. J. (1996). Women and athletic fund raising: Exploring the connection between gender and giving. Journal of Sport Management, 10, 401-416.

Staurowsky, E. J., Parkhouse, B., & Sachs, M. (1996). Developing an instrument to measure athletic donor behavior and motivation. Journal of Sport Management, 10, 262-277.

Stinson, J. L., & Howard, D. R. (2004). Scoreboards vs. mortarboards: Major donor behavior and intercollegiate athletics. Sport Marketing Quarterly, 13(3), 129-140.

Stinson, J. L., & Howard, D. R. (2007). Athletic success and private giving to athletic and academic programs at NCAA institutions. Journal of Sport Management, 21, 235-264.

Stone, K., & Sublett, D. (1992). The UCLA women and philanthropy focus groups. Los Angeles: University of California at Los Angeles.

Strout, E. (2007, July). Courting female donors. Chronicle of Higher Education, 53(44), A21-A22.

Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5th ed.). Boston, MA: Pearson Education.

Taylor, M. A., & Shaw, S. C. (1997). Women as philanthropists: Leading the transformation in major gift fundraising. New Directions for Philanthropic Fundraising, 16, 61-74

Thompson, B. (2004). Exploratory and confirmatory factor analysis: Understanding concepts and applications. Washington, DC: American Psychological Association.

Tsiotsou, R. (1998). Motivations for donations to athletic programs [Electronic version]. The Cyber-Journal of Sport Marketing. Retrieved November 1, 2008, from http://www.ausport.gov.au/fulltext/1998/cjsm/v2n2/tsiotsou22.htm

Tsiotsou, R. (2004). The role of involvement and income in predicting large and small donations to college athletics. International Journal of Sports Marketing & Sponsorship, 6(2), 117-123.

Tsiotsou, R. (2006). Investigating differences between female and male athletic donors. International Journal of Nonprofit and Voluntary Sector Marketing, 11(3), 209-223.

Verner, M. E. (1996). Developing women as financial donors and philanthropists: A way to enhance intercollegiate athletics opportunities. Women in Sport and Physical Activity Journal, 5(1), 29-49.

Wiley, C. E., Shaw, S. M., & Havitz, M. E. (2000). Men's and women's involvement in sports: An examination of the gendered aspects of leisure involvement. Leisure Sciences, 22, 19-31.

Zaichkowsky, J. L. (1985). Measuring the involvement construct. Journal of Consumer Research, 12, 341-352.

Zaichkowsky, J. L. (1994). The personal involvement inventory: Reduction, revision, and application to advertising. Journal of Advertising, 23(4), 59-70.

Zgonc, E. (2010). NCAA: 1981-82--2008-09 NCAA Sports Sponsorship and Participation Rates Report. Indianapolis, IN: The National Collegiate Athletic Association.

Stephen L. Shapiro, PhD, is an assistant professor in the Department of Human Movement Sciences at Old Dominion University. His research interests include college athletic fundraising, ticket pricing, and consumer behavior.

Lynn L. Ridinger, PhD, is an associate professor in the Department of Human Movement Sciences at Old Dominion University. Her research interests include gender equity, consumer behavior, and college athletics.

Table 1.
Profile of Donors Broken Down by Institution

 Institution A Institution B
 (n = 575) (n = 820)

 Male Female Male Female

Gender 68.3% 31.7% 83.2% 16.8%

Ethnicity
White/Caucasian 77.3% 78.4% 94.3% 92.0%
Asian 1.0% 0% 0.6% 0.7%
African American 18.8% 18.2% 2.2% 1.5%
Native American 0.8% 0% 2.5% 5.1%
Hispanic 2.1% 3.4% 0.3% 0.7%

Household Income
Less than $20,000 0.8% 2.4% 1.1% 4.7%
$20,000-$39,999 5.4% 10.7% 3.4% 12.6%
$40,000-$59,999 9.9% 18.3% 11.4% 15.0%
$60,000-$99,999 24.7% 27.2% 16.6% 15.7%
$100,000 or More 59.2% 41.4% 67.5% 52.0%

Marital Status
Single 8.2% 14.9% 5.5% 14.1%
Married 86.4% 60.8% 88.6% 70.4%
Divorced 3.9% 13.3% 4.0% 5.9%
Widowed 0.8% 6.6% 1.0% 8.1%
Separated 0.3% 1.1% 0.6% 0.7%
Other 0.5% 3.3% 0.3% 0.7%

Education
Graduated High School 4.9% 8.3% 2.7% 9.6%
Some College 18.2% 16.7% 16.0% 30.4%
Bachelor's Degree 29.5% 32.2% 37.9% 33.3%
Some Graduate School 11.8% 6.1% 7.7% 5.2%
Graduate Degree 35.6% 36.7% 35.7% 21.5%

Age 56.4 56.1 53.0 52.4

Annual Donation $1,267 $601 $1,399 $928

Donor Length 10.0 8.6 12.0 11.2

 Institution C
 (n = 251)

 Male Female

Gender 84.1% 15.9%

Ethnicity
White/Caucasian 97.6% 95.0%
Asian 0.5% 0%
African American 1.9% 0%
Native American 0% 0%
Hispanic 0% 5.0%

Household Income
Less than $20,000 3.5% 2.6%
$20,000-$39,999 1.0% 7.9%
$40,000-$59,999 9.5% 10.5%
$60,000-$99,999 15.9% 18.4%
$100,000 or More 70.1% 60.5%

Marital Status
Single 10.6% 10.3%
Married 81.6% 82.1%
Divorced 4.8% 2.6%
Widowed 2.9% 5.1%
Separated 0% 0%
Other 0% 0%

Education
Graduated High School 0.5% 7.5%
Some College 4.4% 10.0%
Bachelor's Degree 41.3% 37.5%
Some Graduate School 7.3% 7.5%
Graduate Degree 46.6% 37.5%

Age 50.0 48.5

Annual Donation $1,397 $635

Donor Length 12.2 8.2

Note: Gender, Ethnicity, Household Income, Marital Status,
and Education are Frequency Percentages; Age, Annual
Donation, and Donor Length are Mean Values

Table 2.
Reliability and Validity Scores for the PII

Factors Mean
and Items interitem
 ITTC correlation [varies]

Affective .703 .87
 Exciting .76
 Appealing .79
 Fascinating .72

Cognitive .705 .92
 Needed .81
 Important .81
 Relevant .83
 Means A Lot .80
 Valuable .70

Factors
and Items Factor
 loading AVE SE t

Affective .82
 Exciting .85 -- --
 Appealing .96 .02 51.21 *
 Fascinating .90 .02 46.50 *

Cognitive .84
 Needed .85 -- --
 Important .95 .02 51.42 *
 Relevant .92 .02 46.95 *
 Means A Lot .95 .02 49.36 *
 Valuable .92 .02 47.22 *

Note: * p < .05; ITTC = Item-to-total correlation;
[varies] = Cronbach's alpha coefficient; AVE = Average
variance extracted; SE = Standard error; t = t-values

Table 3.
Donor Mean (Standard Deviation) Gender Comparisons--
Annual Contribution, Donor Longevity, & Age

Variable Male Female

Annual Contribution * $1,360.58(3149.07) $728.76(1235.49)
Donor Longevity * 11.4(9.49) 9.5(8.61)
Age 53.54(12.81) 53.89(11.76)

Note: * = Significant difference;
Bonferonni adjustment significance--p = .017

Table 4.
Donor Gender & Annual Income

Annual Income Level Male Female

Less than $30,000 17 (1.4%) 11 (3.3%)
$30,000-$49,999 44 (3.6%) 37 (11.1%)
$50,000-$69,999 130 (10.6%) 54 (16.2%)
$70,000-$99,999 232 (19%) 73 (21.9%)
More than $100,000 801 (65.4%) 159 (47.6%)

Note: [X.sup.2](4) = 55.07, p < .001