IA Copilot a noté :
Dafotec Récupération De Données

Le 2025-04-26 04:17:00

tigma, Attitudes, and Intentions to Seek Mental Health Services in College Student-Athletes Stigma, Attitudes, and Intentions to Seek Mental Health Services in College Student-Athletes Robert C. Hilliard M.S. West Virginia University, rchilliard@mix.wvu.edu Follow this and additional works at: https://researchrepository.wvu.edu/etd Part of the Counseling Commons, and the Sports Studies Commons Recommended Citation Hilliard, Robert C. M.S., "Stigma, Attitudes, and Intentions to Seek Mental Health Services in College Student-Athletes" (2019). Graduate Theses, Dissertations, and Problem Reports. 4126. https://researchrepository.wvu.edu/etd/4126 This Dissertation is protected by copyright and/or related rights. It has been brought to you by the The Research Repository @ WVU with permission from the rights-holder(s). You are free to use this Dissertation in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you must obtain permission from the rights-holder(s) directly, unless additional rights are indicated by a Creative Commons license in the record and/ or on the work itself. This Dissertation has been accepted for inclusion in WVU Graduate Theses, Dissertations, and Problem Reports collection by an authorized administrator of The Research Repository @ WVU. For more information, please contact researchrepository@mail.wvu.edu. Stigma, attitudes, and intentions to seek mental health services in college student-athletes Robert C. Hilliard Jr. Dissertation submitted to the College of Physical Activity and Sport Sciences at West Virginia University in partial fulfillment of the requirements for the degree of Doctorate of Philosophy in Sport, Exercise, and Performance Psychology Jack C. Watson II, PhD, Chair Samuel Zizzi, EdD Edward Etzel, EdD Aaron Metzger, PhD Department of Sport Sciences Morgantown, WV 2019 Keywords: counseling; help-seeking behavior; well-being; sport psychology; quantitative Copyright 2019 Robert Hilliard ABSTRACT Stigma, attitudes, and intentions to seek mental health services in college student-athletes Robert Hilliard Previous researchers have found several factors that act as barriers to college student-athletes seeking mental health services (López & Levy, 2013; Moore, 2017). One common factor throughout these studies is stigma, which is known to be associated with less favorable attitudes toward seeking help (Moreland et al., 2018). However, researchers have not explored how stigma and attitudes might influence intentions to seek counseling and actual help-seeking behaviors in student-athletes. Additionally, there is a dearth of research identifying the topics for which student-athletes are most willing to seek help. Therefore, the purposes of this study were to investigate predictors of mental health help-seeking as well as identify topics for which college student-athletes are most likely to seek help. The sample consisted of participants (N = 325) from three Division II and III universities. Findings indicated public stigma was significantly related to self-stigma, but social network stigma was not. Self-stigma was related to attitudes and attitudes were related to intentions. Using logistic regression analysis, self-stigma and attitudes were found to be significant predictors of help-seeking behavior. Specifically, both were associated with an increased likelihood of having sought mental health services in the past. Regarding help-seeking topics, drug problems, depression, and excessive alcohol use were the highest rated issues for which student-athletes were likely to seek help, whereas concerns about sexuality, difficulty with friends, and body image were rated the lowest. The results of this study can be used to help sport psychologists and other mental health staff develop programming that might lead to increased service use amongst collegiate student-athletes. Specifically, it appears that using a multifaceted approach to improving attitudes could have the most meaningful effect on encouraging service use. iii Acknowledgements I have saved this section for last because I have struggled to find the words to thank the people in my life who deserve to be in this section. The truth is that the words that are on this page will not ultimately be enough, but I will give it my best shot. My chair, Dr. Jack Watson II, is my obvious starting point. I cannot thank you enough for deciding to take me in four years ago as your student. I have spoken with so many friends across different universities who have had an advisor that has absolutely made or broken their experience in graduate school, and I am lucky enough that I am on the made side. I cannot thank you enough for the opportunities to be involved that you have given me, but also the incredible mentorship that has gone outside of research. Throughout this dissertation process you have inspired me to become a prolific reviewer who mandates everyone uses active voice by writing in first person so that I can begin shifting paradigms. But on a serious note, thank you for the stories, the support, and the confidence to get through this journey. I will be forever grateful. Dr. Sam Zizzi, your influence on my development as a graduate student also cannot be understated. In addition to the guidance up research mountain, your perspectives have constantly challenged me to become a better scholar, practitioner, and person. I will always be thankful for that. Dr. Ed Etzel, you have been such an integral part of my experience at WVU. I am so thankful that you agreed to be part of my committee despite retirement, but again, your role extends far beyond the dissertation. Your continual kindness and compassion is what I will remember the most and I can only hope to one day portray even a sliver of that sort of care with clients and students. Dr. Aaron Metzger, thank you so much for the guidance you have provided over the last year and a half. You are getting a lot of us trickling over into your department and you were always so willing to spend time helping me wade through the mass of confusion that can be statistics, particularly SEM. I also need to thank the rest of the WVU faculty. In some way, each one of you has impacted my life positively and helped me challenge myself to grow. Your support has meant a lot to me and allowed me to enjoy my experience on this path. To all of the WVU program that I have crossed paths with over these four years, thank you. As you all know, completing this journey is no easy task, and it would be substantially more difficult without people around you that you enjoy being with. Thank you for accepting me as I am and for the willingness to provide support whenever necessary. iv I also need to go further down the line with some thanks. Dr. Lindsey Blom and Dr. Robert Batsell, I cannot thank you enough for the mentoring you provided me through my Master’s and undergraduate degrees. Writing this dissertation was a whole lot easier because of the way you challenged me to grow as a writer, and I am forever thankful for that. To my family. Even though you still probably don’t quite understand exactly what I am in school for, you instilled in me a love for education at an early age and always encouraged me to continue finding involvement in things that interested me. Although this journey has continually taken me further and further from home, you are always a part of me. And finally, to my wife, Alisha. I suppose I owe Dr. Blom another acknowledgement of thanks here for bringing us into the same office, but she already knows what she did ☺. Our PhD journeys separated us for far too long, but even in that time, you were a constant source of support and sanity. I can’t wait to start our real journey together, wherever that may take us. v Table of Contents Introduction ............................................................................................................................................... 1 Stigma.......................................................................................................................................... 2 Help-Seeking Concerns ............................................................................................................... 4 The Current Study ....................................................................................................................... 6 Method ...................................................................................................................................................... 7 Participants .................................................................................................................................. 7 Measures...................................................................................................................................... 8 Social network stigma ............................................................................................................. 8 Public stigma. .......................................................................................................................... 9 Self-stigma. .............................................................................................................................. 9 Attitudes................................................................................................................................. 10 Intentions ............................................................................................................................... 10 Help-seeking topics ............................................................................................................... 11 Attention checks .................................................................................................................... 11 Demographics ........................................................................................................................ 12 Procedure ................................................................................................................................... 12 Data Analysis Plan .................................................................................................................... 12 Item parcels............................................................................................................................ 13 Results ..................................................................................................................................................... 14 Data Screening .......................................................................................................................... 14 Preliminary Analyses ................................................................................................................ 15 Relationships Between Stigma, Attitudes, and Intentions ........................................................ 15 Indirect effects. ...................................................................................................................... 16 Predicting Previous Help-Seeking ............................................................................................ 16 Help-Seeking Topics ................................................................................................................. 17 Discussion ............................................................................................................................................... 17 Help-Seeking Intentions and Behavior ..................................................................................... 18 Help-Seeking Topics ................................................................................................................. 22 Implications ............................................................................................................................... 24 Limitations and Future Research............................................................................................... 27 Conclusion .............................................................................................................................................. 29 References ............................................................................................................................................... 30 Appendix A: Figure and Tables............................................................................................................ 39 Figure 1: Structural model. .................................................................................................... 39 Table 1: Sport participation ................................................................................................... 40 Table 2: Correlations ............................................................................................................. 41 vi Table 3: Measurement model ................................................................................................ 42 Table 4: Indirect effects ......................................................................................................... 43 Table 5: Logistic regression .................................................................................................. 44 Table 6: Help seeking topics ................................................................................................. 45 Appendix B: Extended Review of Literature ...................................................................................... 46 The College Population ............................................................................................................. 47 Stigma: An Important Construct Related to Attitudes Toward Help-Seeking .......................... 48 Media Influences on Public Stigma .......................................................................................... 51 Stigma and Help-Seeking Theories ........................................................................................... 53 Measurement of Stigma, Attitudes, and Intentions ................................................................... 55 Attitudes and Intentions to Seek Help in the General Population ............................................. 61 Stigma, Attitudes, and Intentions Toward Help-Seeking in the General Population ................ 65 Mental health interventions to reduce stigma in the general population. .................................. 75 Summary of Stigma and Help-Seeking in the General Population ........................................... 77 Student-Athlete Help-Seeking................................................................................................... 78 Mental Health Interventions ...................................................................................................... 80 Barriers to Help-Seeking for Student-Athletes ......................................................................... 84 Stigma and Attitudes Toward Help Seeking in Athletes........................................................... 87 Likelihood of Seeking Counseling for Specific Issues ............................................................. 93 Future Research ......................................................................................................................... 96 References ................................................................................................................................. 98 Appendix C: Perceptions of Stigmatization by Others for Seeking Help ...................................... 119 Appendix D: Stigma Scale for Receiving Psychological Help ....................................................... 120 Appendix E: Self Stigma of Seeking Help ........................................................................................ 121 Appendix F: Attitudes Toward Seeking Professional Psychological Help Scale-Short Form .... 122 Appendix G: Mental Help Seeking Intention Scale ......................................................................... 123 Appendix H: Help-Seeking Topics .................................................................................................... 124 Appendix I: Demographics ................................................................................................................. 125 Appendix J: Post-hoc analyses for self-stigma ................................................................................. 126 Appendix K: IRB Approval .................................................................................................................... 127 Running head: STIGMA AND COUNSELING INTENTIONS 1 Introduction The recent suicide of a Washington State quarterback sent shockwaves throughout the National Collegiate Athletics Association (NCAA). This is but one of many publicized incidents of mental health issues amongst NCAA athletes, as it is a topic that has been increasing in public awareness (e.g., Associated Press, 2018; Scott, 2018). Its importance is demonstrated through the release of the NCAA Mental Health Best Practices guidelines (NCAA, 2016b) and NCAA research grants that focus on mental health. One component of mental health that has often been studied is stigma associated with seeking counseling services. Although many articles have been devoted to studying stigma and help-seeking attitudes in the general population and even elite athletes, college student-athletes have received significantly less attention. Some authors argue that this population is unique in its needs because of the various cultural factors associated with being a college student-athlete, and therefore are worthy of study (e.g., Etzel, Watson, Visek, & Maniar, 2006; Pinkerton, Hinz, & Barrow, 1989). A myriad of reasons could contribute to the increased need for mental health services amongst student-athletes, such as intense time demands for their sport, highly regimented schedules, and pressures to be successful academically and athletically (Jolly, 2008). Providing evidence for the spread of mental health concerns amongst student-athletes, the results of a recent study at a single university found that mental health issues had an influence on athletic performance in the previous month in 62.6% of the sample (Kern et al., 2017). In a nationwide survey of student-athletes across multiple schools and divisions conducted by the NCAA (2016a), 30% of student-athletes reported feeling overwhelmed in the last month. Gavrilova and Donohue (2018) listed anxiety, depression, substance use, eating disorders, stress, and relationship issues as the top concerns for student-athletes, which mirrors the general college population (Pérez-Rojas et al., 2017). For example, 23.7% of collegiate STIGMA AND COUNSELING INTENTIONS 2 athletes reported clinically significant levels of depression in a recent study (Wolanin, Hong, Marks, Panchoo, & Gross, 2016) and in another study at one university, 14% of the student athletes reported having a mental health condition (Sarac, Sarac, Pedroza, & Borchers, 2018). Despite the clear need for mental health services, several authors argue that student-athletes are still underutilizing these services (e.g., López & Levy, 2013; Moreland, Coxe, & Yang, 2018). A myriad of factors act as barriers to help-seeking for student-athletes. These include lack of time, concerns about confidentiality and being identified in public, worry that clinicians will not understand their lifestyle, and not recognizing the need to seek services (López & Levy, 2013; Moore, 2017; Watson, 2006). In addition to these variables, there is one factor that shows up repeatedly throughout the literature: stigma. As a result, it is likely that stigma plays a significant role in predicting help-seeking intentions and behavior. Stigma and Help-Seeking Corrigan (2004) asserts that two primary types of stigma exist regarding help-seeking, public and self-stigma. Public stigma is external and refers to a belief that society perceives seeking help for mental health treatment as undesirable and individuals who seek help are socially unacceptable. Self-stigma represents an internalization of public stigma in that an individual believes he or she is socially undesirable for seeking treatment. A further delineation of external stigma may also be helpful for understanding the roles that stigma may play in influencing counseling attitudes. Social network stigma is a form of external stigma (Topkaya, Vogel, & Brenner, 2017) that specifically refers to perceived stigma from individuals who are directly within one’s social network (Vogel, Wade, & Ascheman, 2009). Social network stigma is important because an individual may perceive the general public to hold one particular set of beliefs regarding mental health treatment, but the individual might be most concerned about what STIGMA AND COUNSELING INTENTIONS 3 close family and acquaintances think. In one study of Turkish college students that measured public and social network stigma, social network stigma was a unique predictor of self-stigma, although the strength of the pathway was much weaker than for public stigma (Topkaya et al., 2017). Further, there is extensive evidence that these forms of stigma influence attitudes and intentions toward help-seeking in a variety of populations (e.g., Garriott, Raque-Bogdan, Yalango, Schaefer Ziemer, & Utley, 2017). Similarly, stigma is often identified as a barrier to seeking counseling services for college student-athletes (e.g., López & Levy, 2013; Moore, 2017; Moreland et al., 2018). Participants in qualitative studies using athletes across different levels have also reported stigma to be a substantial impediment to seeking counseling services (e.g., Biggin, Burns, & Uphill, 2017; DeLenardo & Lennox Terrion, 2014; Moore, 2017). One reason that stigma might be particularly salient amongst collegiate athletes is the sociocultural aspect of sport, where masculine tenets associated with the sport environment often de-emphasize the need for seeking help and perceive individuals who do so as weak (Gucciardi, Hanton, & Fleming, 2017; Schinke, Stambulova, Si, & Moore, 2018). Therefore, different forms of stigma, including social network stigma, self stigma, and public stigma have all been associated with less favorable attitudes toward seeking counseling (e.g., Bird, Chow, Meir, & Freeman, 2018; Hilliard, Redmond, & Watson, 2018; Ramaeker & Petrie, 2019; Wahto, Swift, & Whipple, 2016). Specifically, social network stigma was found to be associated with increased levels of self-stigma and consequently less positive attitudes in NCAA Division I athletes (Bird, Chow, et al. 2018), whereas public stigma was associated with greater self-stigma and lower attitudes with student-athletes competing across multiple divisions (Hilliard et al., 2018; Wahto et al., 2016). However, no studies have measured public and social network stigma in the same sample of STIGMA AND COUNSELING INTENTIONS 4 student-athletes. Additionally, very little is known about factors that influence intentions to seek professional psychological help in student-athletes. Although attitudes are viewed as a proxy for intentions and behavior (e.g., Ajzen & Fishbein, 1980) and researchers have found that attitudes are related to actual help-seeking behaviors (ten Have et al., 2010), examining actual behaviors would help to get a more accurate picture of student-athlete help-seeking. Recently, some authors have investigated how depression and anxiety symptoms influence help-seeking intentions (Tahtinen & Kristjansdottir, 2018) and actual behavior (Drew & Matthews, 2018). In a recent study of male athletes, attitudes were significantly associated with intentions, although this relationship only accounted for eight percent of the variance (Ramaeker & Petrie, 2019). Additionally, conformity to masculine norms and stigma were indirectly related to intentions. This initial study provides support for the role stigma and attitudes might play in determining intentions. However, given the low variance accounting for intentions in this study, and the lack of an analysis related to actual help-seeking behaviors, more research is needed in this area. Help-Seeking Concerns Research in the general psychology literature has found that specific groups might vary on their willingness to seek counseling for various issues. There has been limited research on this topic amongst student-athletes; however, there is some insight that can be gleaned from this small body of research. One study examined preferences for concerns to discuss in sport psychology consultation amongst athletes from the U.S. and Japan (Naoi, Watson, Deaner, & Sato, 2011). This list included ten options related to performance enhancement as well as various personal concerns. Overall, the top three issues were all performance related: burnout/overtraining, confidence, and dealing with stress. Personal and mental health concerns all ranked in the bottom. Although this makes student-athletes’ willingness to seek help for STIGMA AND COUNSELING INTENTIONS 5 personal concerns appear bleak, the participants were asked to respond to topics they would prefer for sport psychology consultation. They may not have felt that discussing mental health concerns or drug/alcohol issues would be appropriate with a sport psychology consultant, but feel differently about doing so with a mental health professional. However, Moore (2017) recently found of student-athletes that amongst a list of possible behavioral services such as medical, tutoring, and mental health services, mental health related topics received the lowest ratings of comfort. Comfort does not necessarily equate to willingness, as student-athletes are required to meet with academic advisors and sports medicine professionals, so it is understandable that comfort with those services would be higher. Further, when given an option, student-athletes reported a greater willingness to seek help for personal issues from informal sources such as coaches, friends, or family (e.g., Bird, Chow, et al., 2018; Naoi et al., 2011; Maniar, Curry, Sommers-Flanagan, & Walsh, 2001). However, the research on specific presenting concerns for which student-athletes might be interested in seeking counseling is limited. Previous studies have either provided a limited number of sport-related concerns to rate such as overcoming a slump, dealing with injury, or seeking optimal performance (Maniar et al., 2001) or asked participants to rate topics to be discussed during sport psychology consultation (Naoi et al., 2011). Further, research on presenting concerns of college students has found minor differences between genders on why they seek counseling (Pérez-Rojas et al., 2017), but it is unknown if preferences for topics differ between male and female athletes and there has been no discussion of individuals who describe themselves as non-binary. This important line of inquiry can have implications for college counseling centers and sport psychology professionals. Further knowledge of the most and least salient presenting concerns can be used to promote appropriate mental health services and tailor outreach efforts. For example, counseling centers could STIGMA AND COUNSELING INTENTIONS 6 specifically describe the ways they could help student-athletes with their highest rated concerns and destigmatize lower rated concerns that might still be important issues to be addressed with professional help. The Current Study Unlike the broader counseling psychology research investigating other populations, a major limitation of the research with college student-athletes is the lack of studies examining intentions and actual help-seeking behavior, especially when including stigma as a predicting variable. Although researchers have often collected data on previous experience with mental health services, that data has not been included in the analyses. Thus, analyzing the role of stigma in direct relation to attitudes, intentions, and help-seeking behavior would expand understanding of help-seeking factors among student-athletes. Social network stigma may play an important role with student-athletes who have concerns over close others being aware of them seeking counseling (e.g., López & Levy, 2013; Moore, 2017). The measures for public and social network stigma have been found to be able to add unique variance in the prediction of self-stigma. Therefore, understanding the amount of influence of public and social network stigma on help-seeking could help practitioners tailor interventions and outreach to make more student-athletes feel comfortable seeking mental health services. Further, preferred presenting concerns for seeking help are vastly unexplored. The primary aim of this study was to investigate the relationships between public stigma, social network stigma, self-stigma, attitudes, and intentions to seek counseling in college student-athletes. Additionally, the predictive validity of these variables on previous help-seeking was investigated. A secondary aim of this study was to explore willingness to seek counseling for specific topics. STIGMA AND COUNSELING INTENTIONS 7 One framework that can help guide the understanding of health behavior is the Theory of Reasoned Action (TRA; Ajzen & Fishbein, 1980). In the TRA, Ajzen and Fishbein (1980) posit that subjective norms and attitudes combine to form intentions to engage in a behavior, which for the current study was seeking mental health services. Attitudes refer to an individual’s actual beliefs whereas subjective norms refer to the beliefs an individual perceives others to hold. Thus, public and social network stigma are measures that seem to adequately reflect components of TRA. Based on previous research, it was hypothesized that public stigma and social network stigma would both be positively associated with self-stigma. Additionally, it was hypothesized that the relationship between these two external forms of stigma with intentions would be mediated through self-stigma and attitudes. Specifically, it was hypothesized that self-stigma would be negatively associated with attitudes toward counseling, and attitudes would positively relate to intentions to seek counseling. Further, it was hypothesized that all indirect effects in the model would be significant. The current study also explored two different research questions. The first question was: what variables predict if a student-athlete has previously sought mental health services? The second research question was: what are the presenting concerns for which male and female student-athletes are most willing to seek help? Method Participants Participants (N = 328) from a convenience sample of three Division II or III universities in the Mid-Atlantic region of the United States completed surveys. These universities were selected because they did not have sport psychologists on campus, which could influence attitudes toward counseling. Additionally, Division II and III universities tend to have fewer resources available for athletes. The sample was primarily male (n = 224, n = 101 female, 3 did STIGMA AND COUNSELING INTENTIONS 8 not report gender) and predominately from Division III (n = 245, n = 81 Division II, 2 did not report). The student-athletes ranged in age from 18-24 (M = 19.53, SD = 1.29) and contained freshman (n = 112), sophomore (n = 90), junior (n = 73), senior (n = 47), and graduate student (n = 1) athletes. The student-athletes primarily identified as White/Caucasian (84%), followed by Black/African American (7%), biracial (4%), Hispanic/Latino (3%), and less than 1% each of Native American and Asian. Eleven sports were represented, including baseball, basketball, cross country, football, golf, soccer, softball, swimming, track and field, volleyball, and wrestling. Full data related to sport participation can be seen in Table 1. A majority of the sample had no previous experience seeking mental health services (80% no, 19 % yes, 1% did not respond), most had no previous experience with a sport psychology practitioner (88%= no, 10% = yes, 1% did not respond), and only 2% were currently on psychiatric medications. Measures Social network stigma. Perceived stigma from others who are within one’s social network was measured with the Perceptions of Stigmatization by Others for Seeking Help scale (PSOSH; Vogel et al., 2009). Participants responded to five items based on a five-point Likert type scale ranging from 1 (not at all) to 5 (a great deal). The items can be combined for a total score or averaged together for a composite score. Higher scores represent greater perceived stigmatization by others. The internal reliability across five samples including more than 2000 college students was adequate in the validation study (i.e., α = .78 to .89). A reliability of .88 was found in a sample of college student-athletes (Bird, Chow, et al., 2018). In the validation study, the authors found that social network stigma predicted variance in self-stigma beyond other public stigma measures and demonstrated test-retest reliability of r = .77 over three weeks (Vogel et al., 2009). The scale provides the following instructions: “Imagine you had a personal STIGMA AND COUNSELING INTENTIONS 9 difficulty that you could not solve on your own. If you sought counseling services for this issue, to what degree do you believe that the people you interact with would ___”. Following the recommendations of Vogel et al. (2009) to specify the sample, I specified those with whom you interact within the athletic context. A sample item is “think bad things of you.” Cronbach’s α = .82 in this study’s sample. See Appendix C for the scale. Public stigma. The Stigma Scale for Receiving Psychological Help (SSRPH; Komiya, Good, & Sherrod, 2000) was used to measure public stigma. This is a five-item scale with anchors ranging from 0 (strongly disagree) to 3 (strongly agree). Higher scores represent more perceived public stigma. A sample item is “seeing a psychologist for emotional or interpersonal problems carries social stigma.” In the validation study of 311 undergraduate students, the scale demonstrated adequate reliability (α = .72). Other samples using the scale have also found adequate validity above .70 (e.g., Topkaya et al., 2017; Vogel et al., 2009), including .81 in a sample of college student-athletes (Wahto et al., 2016). Komiya et al. (2000) found that a one factor solution fit best and the scale correlated negatively with help-seeking attitudes. In the current study, Cronbach’s α = .64. See Appendix D for the scale. Self-stigma. Self-stigma for seeking treatment was measured using the Self-Stigma of Seeking Help scale (SSOSH; Vogel, Wade, & Haake, 2006). This is a 10-item scale containing five reverse-scored items that is measured on a five-point Likert-scale ranging from 1 (strongly disagree) to 5 (strongly agree). A sample question is “I would feel worse about myself if I could not solve my own problems.” After reverse scoring, higher scores reflect greater levels of self stigma for seeking help. Vogel et al. (2006) validated the scale in several samples of college students, which included more than 1000 participants. The internal consistency across those samples was good (i.e., α = .86 to .91). Additionally, factor analyses confirmed a STIGMA AND COUNSELING INTENTIONS 10 unidimensional structure. Self-stigma predicted counseling attitudes and intentions to seek help above other variables known to influence attitudes and was found to have a two-month test-retest reliability of r = .72. Similar stable psychometric properties have been found in samples of college student-athletes (e.g., Bird, Chow, et al., 2018; Ramaeker & Petrie, 2019). Cronbach’s α = .83 in the current study. See Appendix E for the scale. Attitudes. Attitudes toward seeking counseling was measured using the Attitudes Toward Seeking Professional Psychological Help-Short Form (ATSPPH-SF; Fischer & Farina, 1995). This scale is a shortened version of the original scale. This short version correlated highly with the full version in the initial validation study, r = .87. This scale contains ten items on a Likert-type scale ranging from 0 (disagree) to 3 (agree). Five of the items are reversed scored. After reverse scoring, higher scores indicate more positive attitudes toward counseling. A sample item is “I would want to get psychological help if I were worried or upset for a long period of time.” The scale had a one-month test-rest reliability of r = .80 and had an internal consistency of α = .84, similar to the internal consistency in samples of college athletes (Ramaeker & Petrie, 2019). Cronbach’s α = .78 in this study. See Appendix F for the scale. Intentions. The Mental Help-Seeking Intention Scale (MHSIS; Hammer & Spiker, 2018) was used to measure one’s intentions to seek counseling. The scale was created using a guide on developing Theory of Planned Behavior related instruments, which is an extension of TRA. The MHSIS is a three-item measure that examines intentions using a 7-point Likert-type scale that ranges from 1 to 7. Each of the three questions has different anchors (e.g., extremely unlikely to extremely likely). The three items are averaged together and a higher score represents greater intentions to seek counseling. This three-item version created by Hammer and Spiker was found to be unidimensional and had the greatest predictive validity compared with two other commonly STIGMA AND COUNSELING INTENTIONS 11 used intentions measures. The scale had an internal consistency of α = .94 in the validating study of 405 adults, which was similar to the α = .95 in this study. See Appendix G for the scale. Help-seeking topics. Presenting concerns that student-athletes would be most willing to seek help for were measured using a modified version of the Intentions to Seek Counseling Inventory (ISCI; Cash, Begley, McCown, & Weise, 1975). The original ISCI is a 17-item measure that contains three subscales: academic concerns, psychological and interpersonal concerns, and drug use concerns (Cepeda-Benito & Short, 1998). The four-point Likert-type scale ranges from 1 (very unlikely) to 4 (very likely). Some sample items for which participants indicate they might seek help include loneliness, drug problems, and depression. However, the list omits items that might be relevant to the student-athlete population (e.g., athletic performance). Additionally, some of the language was outdated and might not relate to contemporary college students. Therefore, after conducting a literature search and receiving feedback from licensed psychologists with experience working with collegiate students and student-athletes, some items were modified to reflect recent reports of prominent presenting problems in college counseling centers (Pérez-Rojas et al., 2017) and additional items were added that would be relevant to athletes (e.g., Gavrilova & Donohue, 2018). The final scale contained 20 items. Higher scores indicate a greater willingness to seek help. In this study, Cronbach’s α = .87. See Appendix H for the scale. Attention checks. To reduce the possibility of getting a response set from participants, one attention check question was interspersed throughout the questions as well as a brief distraction scale. Because survey packets were counterbalanced, two separate forms of attention checks were used depending on their place in the packet. One was “please write in the number for ‘a lot’ (4) to answer this question” and the other was “please circle 2 in response to this STIGMA AND COUNSELING INTENTIONS 12 question.” The social identity subscale of the Athletic Identity Measurement Scale (Brewer & Cornelius, 2001) was used as the distractor scale. This is a three-item measure that examines the strength that athletes identify with social aspects of that role. These items include “I consider myself an athlete”, “I have many goals related to sports”, and “most of my friends are athletes.” Demographics. The participants also responded to several demographic questions to provide basic background information. Some of the topics included sport, previous mental health treatment experience, and division of participation. See Appendix I for the questions. Procedure After receiving university IRB approval (Appendix K), the researcher contacted coaches at three universities. If the coach agreed to participate in the study, the researcher met with the team at a convenient time and place to administer the counterbalanced survey packet in person. No coaches or athletic staff were present during data collection. Following data collection, the researcher debriefed participants and they were released. Participants were not compensated and all provided consent. In total, the researcher contacted 26 coaches to participate, and 13 granted permission. Data Analysis Plan A power analysis was conducted prior to collecting data using Soper’s (2013) online calculator for structural equation modeling (SEM). Given the number of latent and observed variables in this study, with a power level of 0.80 and a probability level of .05, the minimum sample size was 268. Therefore, the sample of 312 used for SEM in this study was perceived to be sufficient. To test the hypotheses, AMOS 25.0 (Arbuckle, 2017) was used to examine the data. A two-step model was followed wherein the measurement model was first examined, followed by the structural model (Kline, 2016). This hypothesized model was also compared STIGMA AND COUNSELING INTENTIONS 13 with a plausible alternative model. Finally, mediation through indirect effects via bootstrapping was used to compute bias-corrected percentile confidence intervals, which can be used in cases of multivariate nonnormality (Preacher & Hayes, 2008; Shrout & Bulger, 2002). SEM was used to test the three hypotheses that public and social network stigma will be associated with self stigma, that the relationship of these two stigmas with intentions will be mediated through self stigma and attitudes, and finally that self-stigma will negatively relate to attitudes and attitudes will positively relate to intentions. The goodness-of-fit for the models was assessed using the comparative fit index (CFI), incremental fit index (IFI), root mean square error of approximation (RMSEA), and the standardized root mean residual (SRMR). The guidelines put forth by Hu and Bentler (1999) for good fit were utilized: CFI and IFI ≥ .95, RMSEA ≤ .06, and SRMR ≤ .08. To answer the first research question, a logistic regression was used to test the predictive validity of the variables on previous help-seeking experience. Gender was entered in the first step because it is known to relate to help-seeking, and all three forms of stigma, attitudes, and intentions were entered in step two. Finally, to answer the second research question related to help-seeking topics, descriptive statistics and frequencies were calculated using SPSS. The alpha level for all analyses was set at .05. Item parcels. All scales used in the structural model were unidimensional, which eliminated the possibility of using subscales to represent the latent variable. Thus, to reduce the total number of parameters and improve model fit, parcels were created to represent each latent variable following steps outlined by Russell, Kahn, Spoth, and Altmaier (1998). The scales for public stigma and social network stigma both contained five items, so two parcels were created for each. The scales for self-stigma and attitudes each contained ten items and were given three parcels. Because the intentions scale was only three items, each item was used to represent the STIGMA AND COUNSELING INTENTIONS 14 latent variable. To create necessary parcels, an exploratory factor analysis fit to a one-factor solution was run for the required measures. The factor loadings were balanced on parcels in a way that each parcel would be approximately equal. This was done by alternating highest and lowest loadings on each parcel (Russell et al., 1998). Results Data Screening Initially, 334 surveys were returned out of a possible 336 participants. Six failed the attention check and were subsequently deleted (Spiker & Hammer, 2018). These 328 participants were eligible for the secondary analysis regarding help-seeking topics. This analysis did not rely on statistical assumptions and only required responses to each of the individual items. However, for the main analysis, further cleaning was required. Six additional participants were deleted because they did not complete at least one entire scale, not including the MHSIS. Parent (2013) and Garriot et al. (2017) recommend deleting participants who complete less than 80% of a scale. Parent (2013) further recommended deletion if the overall sample was going to be negligible and under 5%. The participants deleted from the current study did not complete any of the items on a particular scale, falling well under the 80% range. Additionally, Hammer and Spiker (2018) state that a mean should not be calculated for individuals who do not complete all three items on the MHSIS, which resulted in eight deletions. This left 314 participants to examine missingness and statistical assumptions. Overall, there were 16 missing values out of a total of 10,362 values. Percentage of missingness ranged from 0.3-1.3%. Due to such low missingness, values were replaced using expectation maximization (Schlomer, Bauman, & Card, 2010). Next, steps outlined by Tabachnick and Fidell (2013) were followed for examining outliers and assumptions. All variables appeared to be normally distributed, although the STIGMA AND COUNSELING INTENTIONS 15 skewness ratio for social network stigma exceeded 3.29, suggesting possible nonnormality (Tabachnick & Fidell, 2013). However, according to Byrne (2016), extreme kurtosis is more concerning for SEM, and the skewness values were not above the threshold indicating problematic nonnormality. As a result, to ease interpretation, the variable was left untransformed. There was one univariate outlier with a z-score above 3.29 that was deleted, and one multivariate outlier with a Mahalanobis distance value exceeding the critical value for the degrees of freedom in this study, and was subsequently deleted. This meant that 312 participants were included in the SEM and logistic regression analyses. The data met all other multivariate assumptions. The last step was to check for multivariate normality, and Mardia’s coefficient was 9.46, suggesting mild multivariate nonnormality (Byrne, 2016). Preliminary Analyses Means, standard deviations, and correlations were examined between the variables included in the SEM and logistic regression analyses. These values can be seen in Table 2. The correlations between all variables were significant and in the expected direction. The strongest relationship was between attitudes and intentions, r = .61, p < .001, and the weakest was between social network stigma and intentions, r = -.12, p = .04. Relationships Between Stigma, Attitudes, and Intentions The measurement model provided good fit to the data, χ2(55, N = 312) = 109.19, p < .01, CFI = .98, RMSEA = .05 [90% CI = .04, .07], IFI = .98, SRMR = .04. All items loaded significantly on their observed factors, with loadings ranging from β = .56 to .97. Table 3 displays the full values from the measurement model. Overall, the goodness-of-fit and the loadings suggested that the latent variables were adequately measured. STIGMA AND COUNSELING INTENTIONS 16 The structural model was examined next. The researcher first tested the hypothesized model with public and social network stigma as exogenous variables whose relationship with intentions was mediated through self-stigma and attitudes. This model demonstrated good fit to the data, χ2(60, N = 312) = 119.75, p < .01, CFI = .98, RMSEA = .06 [90% CI = .04, .07], IFI = .98, SRMR = .05. This model accounted for 47% of the variance in self-stigma, 41% of the variance in attitudes, and 50% of the variance in intentions. This model can be seen in Figure 1. Of note, all hypothesized pathways were significant, except for the one between social network stigma and self-stigma (β = -.02, p = .83). Next, a plausible alternative model (Kline, 2016) was tested, where public and social network stigma each had a direct path to intentions. This partially mediated model also demonstrated good fit, χ2(58, N = 312) = 118.34, p < .01, CFI = .98, RMSEA = .06 [90% CI = .04, .07], IFI = .98, SRMR = .05. The original pathways all remained significant, but the two new direct pathways from public and social network stigma to intentions were not significant. A test to compare the two models found that they fit the data equally, χ2(2, N = 312) = 1.41, p = .49. Given that the additional pathways were not significant, the original, smaller parsimonious model was utilized to test indirect effects. Indirect effects. To calculate indirect effects, 5000 bias-corrected bootstrap samples were requested using AMOS. As shown in Table 4, all indirect effects in the model were significant with the exception of any pathways that included social network stigma. In addition to the AMOS output for significance, the 95% confidence intervals for the pathways deemed significant did not contain zero, confirming that they are significant (Shrout & Bulger, 2002). Predicting Previous Help-Seeking STIGMA AND COUNSELING INTENTIONS 17 To answer the first research question, the researcher conducted a binary logistic regression using the measured variables as predictors of previous help-seeking history. An additional assumption for logistic regression includes testing the linearity of the logit, and this assumption was met. Results from this regression analysis are seen in Table 5. The model was not significant after only including gender in step one, but the full model was significant, χ2(6, N = 310) = 28.19, p < .001. The Hosmer-Lemeshow test for this model indicated good fit, χ2(8) = 5.63, p = .69. As seen in the table, self-stigma (Expβ = 2.28, p = .01) and attitudes (Expβ = 3.76, p = .001) significantly predicted previous mental health service use. These odds ratios both suggest that as the individual scored higher in self-stigma or attitudes, they were more likely to have previously sought mental health services. Help-Seeking Topics To answer the second research question, the researcher compiled means and standard deviations for each variable on the ISCI. Overall, drug problems, depression, and excessive alcohol use were the highest rated issues for which student-athletes were likely to seek help. The lowest rated topics were concerns about sexuality, difficulty with friends, and body image. When split by gender, the results remained primarily the same. For women, the highest ranked topics were anxiety, depression, and drug problems, whereas the lowest ranked were concerns about sexuality, difficulty with friends, and difficulty with sleeping. For men, the highest and lowest rated topics were identical to the overall sample, although the order varied for the lowest ranked. Body image was the second lowest ranked topic, followed by difficulty with friends at third. The results for all 20 items on this scale can be seen in Table 6. Discussion STIGMA AND COUNSELING INTENTIONS 18 The purpose of the current study was to explore the role of various stigmas on intentions to seek counseling and how these factors were associated with previous help-seeking experience. An additional aim of this study was to determine which topics student-athletes are most and least willing to seek help for professionally. The researcher hypothesized that public stigma and social network stigma would be positively associated with self-stigma. In turn, self-stigma would be negatively associated with attitudes toward counseling, and attitudes would then be positively associated with intentions to seek counseling. Further, it was hypothesized that all indirect effects would be significant. There were no hypotheses for the research questions. Overall, the hypotheses were mostly supported. Help-Seeking Intentions and Behavior Previous researchers have identified stigma as having a strong association with attitudes toward help-seeking in college athletes (Bird, Chow, et al., 2018; Wahto et al., 2016). Specifically, public stigma, social network stigma, and self-stigma have all been linked with counseling attitudes, with self-stigma having the strongest link (Bird, Chow, et al., 2018; Hilliard et al., 2018; Wahto et al., 2016). In the current study, social network stigma was not significantly associated with self-stigma. The overall low scores and restricted range of responses of social network stigma may have contributed to its lack of significance. Alternatively, it is possible that contemporary athletes receive more positive messages from teammates and coaches related to mental health help-seeking than in the past and thus these beliefs are not being internalized in the form of self-stigma. However, public stigma, self stigma, and attitudes were related to intentions. This aligns with the findings of a recent study of male college student-athletes, where self-stigma and attitudes were associated with intentions (Ramaeker & Petrie, 2019). This suggests that those with more positive attitudes toward seeking STIGMA AND COUNSELING INTENTIONS 19 help will also endorse greater intentions to seek mental health counseling if necessary. When comparing the overall levels of stigma and attitudes to other studies of student-athletes, there are minimal differences (Bird, Chow, et al., 2018; Hilliard et al., 2018; Wahto et al., 2016). The largest difference between means on a measure was 0.21 for public stigma on a 5-point scale (Hilliard et al., 2018). Thus, when interpreting these results within the broader literature, the current sample does not appear to be different from other study populations. These findings support the assertions of the TRA (Ajzen & Fishbein, 1980) that more positive attitudes toward a health behavior will result in increased intentions of engaging in that behavior. Subjective norms is another component of the TRA and was measured via public stigma. Public stigma contributed to inform intentions, as seen through the indirect relationship between public stigma and intentions through mediating pathways, including self-stigma, an important variable related to attitudes. The importance of these variables in this theoretical construct was further established when examining predictors of previous help-seeking behavior. Similar to the structural model, more positive attitudes were indicative of the greatest increase in likelihood that an individual had previously sought mental health services. Attitudes have previously been associated with actual service use (ten Have et al., 2010), and in this sample they were the strongest predictor of previous help seeking. This finding is slightly inconsistent with the TRA, where the most proximal measure of behavior (i.e., intentions) should be the strongest predictor. Additionally, there was a strong correlation between attitudes and intentions present in the study. Given this information, future researchers may want to examine the utility of intentions over and above attitudes for examining student-athlete help-seeking behavior. One possibility for why attitudes may have been more important than intentions may be the STIGMA AND COUNSELING INTENTIONS 20 unpredictability associated with intentions. Half of the variance in intentions was accounted for in this study. In the only other study of student-athletes to measure intentions in line with this framework, 38% of the variance in attitudes was accounted for, but only 8% for intentions (Ramaeker & Petrie, 2019). Part of this low variance may be due to the measure of intentions that was used by Ramaeker and Petrie, which is arguably not a good measure of actual help seeking intentions (Hammer & Spiker, 2018). It is difficult for an individual to predict how one might act in a certain situation, which is what intention measures ask participants to do. Given the low number of individuals who were currently seeking help and the scores on the various measures, it appears that the sample in the current study was at fairly low risk. This may have contributed to ambivalent intentions scores. Attitudinal measures, on the other hand, probe what the participant currently believes. This is not to say that intentions measures have no use in examining help-seeking behavior, as multiple meta-analyses have found that they are relatively accurate predictors of behavior (e.g., Armitage & Conner, 2001; Sheppard, Hartwick, & Warshaw, 1988). In fact, in a recent sample of college students, intention was found to predict previous mental health service use, although attitudes were not included in the logistic regression so a comparison could not be made (Li, Denson, & Dorstyn, 2018). In their study, the odds ratio for intentions varied between 1.03-1.04 across various blocks, indicating a fairly weak influence. However, the decision to seek help is often not a fully rational or linear process, and therefore intentions may not always be a strong predictor future behavior. Thus, future research using variables to measure the full TRA model and using the MHSIS as the intentions measure would clarify the relationship between intentions and help-seeking behavior. An unusual finding in this study was that self-stigma was associated with an increased likelihood of having sought services in the past. This is a novel finding that contrasts previous STIGMA AND COUNSELING INTENTIONS 21 studies of help-seeking. It is unknown why this occurred in the present sample, but there are a few potential statistical and practical explanations for this relationship. Levels of self-stigma did not vary between groups of individuals who had or had not previously sought counseling and there was no correlation between self-stigma and previous help-seeking. Further, in post-hoc logistic regressions to explore this finding, self-stigma was only a significant predictor when coupled with attitudes (See Appendix J). Thus, it is possible that attitudes were having a statistical influence on self-stigma and this finding was an anomaly. However, this finding supports the assertions of Schomerus and Angermeyer (2008) that the relationship between attitudes, self-stigma, and help-seeking behavior might not be as clear as previously believed. Regardless, further research is needed to corroborate this result. Practically, one possible explanation is that self-stigma was developed as a response to seeking treatment. No research was located that temporally examined the development of self stigma, and it is possible that negative experiences with a counselor led to increased levels of self-stigma. A second plausible explanation is that the levels of self-stigma was overpowered by other needs of the individuals. Previous researchers have found that individuals who recognize they have a mental health issue or concern are more likely to endorse greater levels of self stigma (e.g., Corrigan & Watson, 2002; Jones, Keeling, Thandi, & Greenberg, 2015). Further, some individuals reject the meaning of the self-stigma, even if they possess it (Corrigan & Watson, 2002). Therefore, it is possible that the student-athletes in the current study who had previously sought mental health services were aware of their concerns, which led to increased levels of self-stigma. However, as student-athletes, they might have also recognized that their mental health influences their performance. A previous study has confirmed this as one reason why student-athletes sought mental health services (Bird, Chow, & Cooper, 2018). With the STIGMA AND COUNSELING INTENTIONS 22 desire to be as healthy as possible to compete at their highest level, their athletic identity or competitive desires may have played a more prominent role in the decision to seek help, negating the negative effects of self-stigma. However, this is currently speculation. Future examination of the temporal associations of self-stigma and other help-seeking related variables would clarify this relationship. Help-Seeking Topics The second research question was designed to assess the topics for which student-athletes report being most willing to seek help, ranking them from most to least likely. The results indicated that overall, student-athletes would be most likely to seek help for drug problems, depression, and excessive alcohol use and least likely to seek help for concerns regarding sexuality, difficulties with friends, and body image. These results were similar across gender as well, except that anxiety was rated as highest for women, eliminating excessive alcohol use, and difficulty sleeping was the lowest rated, eliminating body image concerns. These results fall in stark contrast to the preferences of the student-athletes in Naoi et al. (2011). In their study, mental health concerns and alcohol and drug issues ranked ninth and tenth, respectively, out of a list of ten items. However, the differences in methodology explain this inconsistency. The participants in Naoi et al. (2011) were asked to rate preferences for topics when meeting with a sport psychology consultant. In the current study, participants responded to a list of topics for which people often seek counseling. Therefore, it was apparent that these issues would be discussed with a mental health professional and not a sport psychology consultant. This aligns with previous research that has found that student-athletes display preferences for professionals based on match regarding their particular concern (Maniar et al., 2001). Thus, it makes sense that the student-athletes reported high levels of willingness to seek STIGMA AND COUNSELING INTENTIONS 23 help for drug and alcohol use and depression with a counselor, but did not prefer to discuss these issues with a sport psychology consultant (Naoi et al., 2011). Interestingly, seeking counseling for athletic performance was rated fairly low in this sample, which further indicates that the match with the professional might be an important indicator of their intentions to seek help for a particular issue. The issues rated most highly in this study have also been identified as important mental health concerns that student-athletes face (Gavrilova & Donohue, 2018). These findings also broadly mirror the results from a large-scale study of presenting concerns in college counseling centers, where the top three concerns for athletes were anxiety, stress, and depression (Pérez Rojas et al., 2017). However, alcohol and drug use was not included in their ranking system, so it is not known how many athletes presented with those concerns. Nevertheless, the high willingness to seek help for drug or alcohol issues is a promising step forward, as these are often deeply stigmatized issues. One possible explanation for the high rankings for these topics could be that the majority of the sample came from a Christian affiliated school with a strict drug and alcohol use policy. This environmental factor may have led to a greater willingness to seek help for those particular topics. Comparisons between the help-seeking topics scale and other student athlete populations cannot be made. Equally important to the highest rated topics are those that were ranked low amongst the athlete group. Although student-athletes could benefit from seeking mental health services for all items on the scale, the reasons for these three being lo