Test and study worry and emotionality in the prediction of college students' reasons for drinking: an exploratory investigation.

Abstract

Educational environments can place an exorbitant strain on the psychological well-being of students, and oftentimes students resort to substance use as a means to escape the stress of performance expectations. This study explored the influence of test and study worry and emotionality on students' reported reasons for consuming alcohol. The Reasons for Drinking Scale (RFD; Cronin, 1997) and Test and Study Attitudes Inventory (TSAI; Spielberger & Lunsford, 2001) were used to assess student reported reasons for consumption and test and study worry and emotionality, respectively. Results indicated that students who experienced test and study related worry and emotionality reported drinking for tension reduction more so than for social camaraderie or mood enhancement. Further, women reported that mood enhancement and tension reduction were less likely to be reasons why they consume alcohol. These findings suggest that strategies to reduce test and study anxiety may also decrease alcohol consumption related to tension reduction.

KEYWORDS: Drinking Behavior, Alcohol, Student Drinking, Worry, Emotionality, College Students, Test Anxiety, Test Performance

INTRODUCTION

Alcohol use among college students is a mounting concern, and myriad problems can arise from frequent and excessive alcohol consumption including sexual promiscuity or assault, decrements in academic grade point average and possible academic suspension or expulsion, incursions with authorities and possible arrest, and even physical disability or death (Camatta & Nagoshi, 1995; Park, 2004). Although heavy drinkers often consume alcohol to become inebriated, the vast majority of college students drink for a variety of other reasons, both positive and negative. Park (2004) recently reported that although students seem to understand the negative consequences of drinking prior to consumption, they often engage in the behavior because of the frequent positive benefits of consuming alcohol. Further, she noted that increased consumption, although linked to more serious negative effects of use, was also reportedly linked to much higher expectations of extremely positive consequences of drinking including meeting new friends, celebrating, relieving stress, increasing social lubrication, and having romantic encounters. Similarly, Yokoyama, Nishikitani, and Araki (1999) reported that students frequently engage in drinking for two predominant reasons, either to forget a bad mood and escape difficult life circumstances or to increase sociability with peers and members of the same or opposite sex. Students less often, according to these authors, engaged in drinking to experience a heightened sensation or because of perceived peer pressure.

The Influence of the Self-Medication Hypothesis on Consumption

One hypothesized reason for the insurgence of student drinking is the reduction of psychological symptoms. Novak, Burgess, Clark, Zvolensky, and Brown (2003) found that sensitivity to anxiety was not related directly to the amount of alcohol consumed but was related to students' motivations for engaging in substance using behavior, with men more likely to consume alcohol for tension reduction purposes than women. Swendsen, Tennen, Carney, Affleck, Willard, and Hromi (2000) used the experience sampling method to link nervous behavior earlier in the day to increases in alcohol consumption later in the day and again found that men were more likely to assuage anxiety with alcohol than were women. The phenomenon of attempting to reduce psychological symptomatology through substance use is often referred to as the "self-medication hypothesis." Proponents of the self-medication hypothesis have argued that individuals seek to reduce psychological symptoms with excessive substance consumption, a coping mechanism that allows these individuals to function in the face of perceived adversity. This can be especially true in the case of anxiety symptoms, as those who suffer from anxiety disorders often report drinking excessive quantities of alcohol because of the life stressors they experience.

However, support for the self-medication hypothesis as it relates to anxiety disorders is certainly not universal, and researchers have proposed alternative explanations for the frequent co-occurrence of anxiety symptoms and alcohol use. Schuckit and Hesselbrock (1996), among others, have argued that anxiety symptoms may precede those of alcohol abuse because the former tend to occur at a much earlier age and have not been conclusively shown to cause excessive alcohol consumption. Although research by Christie, Burke, Regier, Rae, Boyd, and Locke (1988) reported that psychological distress resulted in increased likelihood of substance use in their study, other research by Mensch and Kandel (1988) and Hansell and White (1991) found no direct influences of sustained psychological distress on substance abuse patterns.

Interestingly, other recent research on the self-medication hypothesis (e.g., Degenhardt & Hall, 2001) has argued that sustained periods of heavy substance use may serve to increase psychological symptoms rather than decreasing them regardless of the motivations for initiating consumption. In adolescents and college-aged students, the mechanism at work may involve association with deviant peers. As young adults experience psychological difficulties, they may lose the motivation to conform to societal expectations and may search for ways to increase intrapsychic satisfaction with the self. This search may result in "acting out behavior" including the execution of deviant acts or association with deviant peers, both which place the individual at greater risk for using alcohol and drugs (Bryant & Zimmerman, 2002; Damphousse & Kaplan, 1998). As psychological symptoms worsen, substance use may increase in an effort to reduce the unpleasant and unwanted effects of the disorder that led the young person to initially consume.

Research on the Relationship Between Reasons for Drinking and Anxiety

Despite disparity in the research literature on the origins of the frequently found co-occurrence of anxiety and substance use disorders, there are numerous studies that have linked periods of stress to increased consumption of alcohol in several populations including college students and psychiatric patients. The vast majority of these studies have found that heavy drinkers report alcohol consumption for reasons of negative reinforcement (i.e., to remove unwanted psychological effects) whereas light or social drinkers tend to consume alcohol for reasons of positive reinforcement (i.e., to be more sociable and fun). Bergman and Harris (1985) reported that among a small sample of young adult psychiatric patients (n = 65), the primary reason for consuming alcohol was to relieve the anxiety associated with interacting with others. Similarly, Test, Wallisch, Allness, and Ripp (1989) found that schizophrenic patients frequently consumed alcohol to relieve both boredom and anxiety symptoms. Similarly, in a sample of psychiatric outpatients, Carey and Carey (1995) found that those individuals who drank most heavily reported doing so to obtain relief from their psychological symptoms including nervousness and anxiety. And Mehlum (1999) recently surveyed 1,373 Norwegian men and women who served peacekeeping missions in Lebanon during the late 1970s through the early 1990s on both the amount of alcohol consumed and the respondents' reasons for consuming alcohol following particularly traumatic events during their tours of duty. Results indicated that only a minority of individuals reported using alcohol to reduce anxiety and tension, relieve restlessness or to abate the effects of psychological stress. However, among those individuals experiencing the highest levels of stress (e.g., those who witnessed the death of a colleague or atrocities committed against civilians), alcohol consumption increased substantially "during the peacekeeping missions and was sustained after extrication from the stressful situation.

Similarly, a number of previous studies have found that college students consume alcohol to reduce tension, relax, and escape from the stresses of college life (Yokoyama, Nishikitanl, & Araki, 1999). One explanation for such behavior is Tension Reduction Theory (TRT) that states that college students consume alcohol to reduce tension because of alcohol's known depressive effects. Kalodner, Delucia, and Ursprung (1989) found that highly anxious college students drank substantially more (twice as much in men and 1.5 times more in women) than a comparable sample of low anxiety students. Yokoyama, Nishikitani, and Araki (1999) reported that problem drinking in Japanese pre-college age students (ages 15-18 years old) was related to escapism and tension reduction. And Miller, Miller, Verhegge, Linville, and Pumariega (2002) recently found that college athletes who complained of psychological difficulties, including anxiety, were much more likely to abuse alcohol and engage in risk-taking behavior than those individuals not experiencing psychological difficulties.

However, research has shown that students consume alcohol for a variety of reasons and not just to reduce tension and relieve stress. Typically, both personal and social motivations are identified as reasons why students drink, and these often include socializing with others and mood enhancement (Billingham, Parillo, & Gross, 1993; Haden & Edmundson, 1991; Johnston, O'Malley, & Bachman, 1989). Personal motivations such as mood enhancement have been shown to predict problem drinking (Billingham et al., 1993; Wood, Nagoshi, & Dennis, 1992), daily alcohol use (Johnston, O'Malley, & Bachman, 1989) and heavy drinking (Cutter & O'Farrell, 1984). Social motivations, particularly among college students, have been found to predict frequent, non-problematic drinking (Haden & Edmundson, 1991; Klein, 1992; Wood et al., 1992). Further, it has been suggested that non-problematic, frequent alcohol use among college students may be related more to social norms than to attempts to reduce psychological symptoms (Wood et al.).

In sum, past research has linked feelings of anxiety and a need to relieve such feelings to reasons for alcohol consumption. Theories such as the self-medication hypothesis and tension reduction theory have attempted to explain this relationship. However, support for such notions is not universal, and other lines of research have suggested that many individuals with psychological difficulties do not use substances to cope with life difficulties, and that frequent use of alcohol as a coping mechanism for psychological difficulties can actually exacerbate psychological symptoms rather than having the intended effect of reducing them.

Test and Study Worry and Emotionality: Predictors of Substance Use?

Anxiety has been found to adversely affect myriad aspects of college student performance including both preparation for and performance on measures of academic achievement (Sud & Prabha, 1995), but those students suffering from academic performance-related anxiety often experience other problems as well. Recent research has shown that students who suffer from test anxiety may also experience lower perceptions of familial communication, encouragement of personal growth, and familial system maintenance (Peleg-Popko, 2002), may demonstrate immune system deficiencies that lead to higher rates of illness than in comparatively healthy individuals (Borella, Bargellinin, Rovesti, Pinelli, Vivoli, Solfrini, et al., 1999; Marshall, Agarwal, Lloyd, Cohen, Henniger, & Morris, 1998; Wadee, Kuschke, Kometz, & Berk, 2001), and may be prone to negative health related factors later in life if their anxiety symptoms are sustained (Stanley, Now, Bourland, Beck, & Averill, 2001). Thus, anxiety about academic performance has been linked to various factors that can inhibit student achievement, motivation, and self-efficacy.

Test anxiety in college students has often been decomposed into two main elements: (a) worry, or the cognitive rumination about aspects of school-related subjects, including performance on tests, quizzes, and homework; and (b) emotionality, or the physiological expression of anxiety about school-related subjects. Research has examined the independence of these constructs, the majority of which has indicated the unique contributions of both worry and emotionality to the understanding of the more complex construct of test anxiety (Williams, 1996). Past research has also indicated that worry is often a stronger negative predictor of test performance than emotionality (Harris & Cumming, 2003; McIlroy, Bunting, & Adamson, 2000) and that women tend to demonstrate higher levels of test worry and emotionality than male counterparts (Crocker, Schmitt, & Tang, 1988; Williams, 1996).

Purpose of the Present Study

Given the state of current educational environment in which students often report increased pressure to perform well, it is likely that college students will not only experience test and academic performance-related anxiety, but that these same students may resort to substance use as a means of escaping these stress-inducing situations. Although general anxiety symptoms (independent of test anxiety) have been linked to various reasons for drinking in college students, it is interesting to note that little concerted attention as been directed toward examining the relationship between test and study anxiety and student reported reasons for alcohol consumption. Further, little concerted research has been directed toward examining whether reasons for drinking change during students' years in college. Indeed, freshmen have fewer social support resources at their disposal than students who have been in the college environment for several years (Klein, 1992).

Thus, the purpose of the present study was to explore the relationship between test and study worry and emotionality and student reported reasons for consuming alcoholic beverages. We used the Reasons for Drinking Scale (RFD; Cronin, 1997) to assess student reported reasons for alcohol consumption and the Test and Study Attitudes Inventory (TSAI; Spielberger & Lunsford, 2001) to assess both test and study worry and emotionality. Specifically, we developed three testable hypotheses to explore the objectives of the present study based on the previously presented research:

(a) Students who reported experiencing test and study related worry and emotionality were predicted to score higher on the Tension Reduction subscale of the RFD;

(b) Men were predicted to report tension reduction as a stronger reason to drink than women, with women predicted to exhibit higher levels of test and study worry and emotionality than men; and

(c) Underclass students were predicted to report drinking for reasons of tension reduction and social camaraderie more often than upperclass students.

METHOD

Participants

Participants in the present study were 365 students enrolled at a small, regional, liberal arts college in the Southeast. The mean age of the sample was 20.57 years old (SD = 4.84), with 67.7% (n = 247) of the participants being female. Approximately 45.8% (n = 167) of the participants were White, 12.3% (n = 45) were Black, 11.0% (n = 40) were Latino/Latina, 13.7% (n = 50) were Asian/Pacific Islander with the remaining 17.2% (n = 63) self-labeling themselves as "other." Approximately 91.5% (n = 334) of the participants reported never being married, and 48.5% (n = 177) of the respondents reported living in campus housing. Average college GPA was 3.20 (SD = .48) with average reported Scholastic Aptitude Test (SAT) scores of 1049.86 (SD = 125.43).

Measures

Reasons for Drinking Scale (RFD; Cronin, 1997). The RFD is a 25 item, three point Likert-type instrument that is purported to measure the reasons that individuals consume alcohol. Responses to individual items range from 0 (not a reason) to 2 (a major reason). Factor analytic studies have identified three subscales, with higher scores on those subscales indicating a stronger reason for drinking: Mood Enhancement (8 items), Social Camaraderie (8 items), and Tension Reduction (9 items). Cronin (1997) reported score reliabilities ranging from .79 (Social Camaraderie) to .86 (Mood Enhancement). Coefficient alphas for the RFD subscales in the present study indicated acceptable internal consistency reliability, with alphas of .89, .87, and .87 for the Social Camaraderie, Mood Enhancement, and Tension Reduction subscales, respectively.

Test and Study Attitudes Inventory (TSAI; Spielberger & Lunsford, 2001). The TSAI is a 32 item, 4-point Likert-type instrument that is purported to measure both test and study worry and emotionality. Item responses range from 1 (almost never) to 4 (almost always). Factor analytic studies of the TSAI have indicated a four factor structure, with higher scores on the subscales indicating higher levels of worry or emotionality: (a) Test Worry (8 items); Co) Test Emotionality (8 items); (c) Study Worry (8 items); and (d) Study Emotionality (8 items). Coefficient alphas for the TSAI subscales in the present study indicated acceptable internal consistency reliability, with alphas of .87, .85, .83, and .83 for the subscales of Test Worry, Test Emotionality, Study Worry, and Study Emotionality, respectively.

Procedure

Study participants were recruited through undergraduate level courses and had the opportunity to participate in the research during their normal class time for extra credit toward their final grade. Students were informed about the purpose of the study, their voluntary participation in it, and the conditions under which they would receive credit for participation. No students declined to participate.

RESULTS

Prior to conducting the substantive statistical analyses, the data were examined for multivariate normality. It has been suggested that failure to examine multivariate normality could result in spuriously accentuated or attenuated results (Fan, 1996). To assist in evaluating the multivariate normality of the variables, a program called MULTINOR was used (Thompson, 1990). This procedure involves first computing the Mahalanobis distance of each case from the centroid of all group means. Cases that have very large Mahalanobis ([D.sup.2]) distances are considered multivariate outliers and are typically removed from further analysis. Based on this approach, the Mahalanobis distances were computed for each of the 365 cases on the three subscale of the RFD and the four subscales of the TSAI. None of the cases met criteria for removal.

Canonical Correlation Analysis

To test the first hypothesis, a canonical correlation analysis (CCA) was conducted using the four subscales of the TSAI as predictors of the three subscales of the RFD to evaluate the multivariate shared relationship between the two variable sets of test and study worry and emotionality and students' reasons for drinking. The analysis yielded three functions with squared canonical correlations ([R.sub.c.sup.2]) of .068, .034, and .004 for each of the three successive functions. Collectively, the full model across all functions was statistically significant, Wilks' A = .897, F (12, 674.96) = 2.362, p = .006. Because Wilks' A represents the variance unexplained by the model, (1 - A) yields the full model effect size in a r-squared metric. Thus, for the set of three canonical functions, the r-squared type effect size was .103 indicating that the model can explain 10.3% of the variance shared between the variable sets.

A dimension reduction analysis allowed testing the hierarchal arrangement of functions for statistical significance. As noted, the full model (functions 1-3) was statistically significant. Function 2-3 and 3-3 were not statistically significant, F (6, 512) = 1.638, p = .135 and F (2, 257) = .466, p = .628, respectively. Given the [R.sub.c.sup.2] effects for each function, only the first function was considered noteworthy in the context of the current study (10.3% of shared variance between variable sets). The last two functions (2-3 and 3-3) each explained 3.7% and 0.3%, respectively, of the variance in variable sets after extraction of prior functions.

Table 1 presents the standardized canonical function coefficients and structure coefficients for function 1. The squared structure coefficients are also provided. Examination of the function 1 coefficients revealed that the relevant predictor variables were primarily Test Emotionality (1.288), Test Worry (-.988) and Study Worry (.957). This conclusion was also supported by the squared structure coefficients, as these predictor variables also had the larger canonical function coefficients. A slight exception involved the Test Worry subscale that had a large function coefficient but a much smaller structure coefficient. This result is likely due to the multicoliinearity that these two variables have with the other criterion variables, and all variables' structure coefficients have the same sign, indicating that they are all positively related.

Regarding the criterion variable set in function 1, only Tension Reduction was the primary contributor to the criterion synthetic variable. This conclusion is supported upon examination of the structure coefficients, as Tension Reduction had the largest function and structure coefficient. However, both Social Camaraderie and Mood Enhancement exhibited large structure coefficients but small function coefficients. Again, this result is likely due to the multicollinearity of these two variables with the other criterion variable.

Differences Between Men and Women on RFD and TSAI Scores

To test hypotheses two and three, additional statistical analyses were conducted to determine if differences in the subscale scores were influenced by gender or year in college. We conducted a two-way multivariate analysis of variance (MANOVA) with gender and year in school as the fixed factors and the three subscales of the RFD and the four subscales of the TSAI as the dependent variables. Table 2 presents means and standard deviations for the RFD and TSAI subscales by gender and year in college. Also presented in Table 2 are Cohen's ds (Cohen, 1988) between men and women on each of the subscales.

The gender by year in college interaction was not statistically significant, Wilks' A = .943, F(21,606.42) = .596,p = .923, and only accounted for 5.7% of the variation in the dependent variables. Similarly, the main effect of year in college was not statistically significant, Wilks' A = .930, F(21, 606.42) = .739, p = .794, and only accounted for 7.0% of the dependent variable variance. However, the main effect of gender was statistically significant, Wilks' A = .895, F(7, 211.00) = 3.53,p = .001, and accounted for 10.5% of the variance in the dependent measures. Post-hoc analysis of variance (ANOVA) indicated statistically significant differences primarily on the Mood Enhancement (p = .0004) and Tension Reduction (p = .007) subscales of the RFD. Examination of the means presented in Table 2 indeed indicated that women reported Mood Enhancement and Tension Reduction to be a lesser reason why they drink across the four college student groups. When examining differences between men and women regardless of year in college on standardized mean differences (i.e., Cohen's d), all of the differences between men and woman would be considered noteworthy but small effects (i.e., Cohen's d greater than .20) with the exception of Study Worry (d = -.382), Mood Enhancement (d = .576) and Tension Reduction (d = .447) which would be considered medium effects.

DISCUSSION

The purpose of the present study was to determine if test and study worry and emotionality predicted student reported reasons for consuming alcohol. It was hypothesized that (a) students who experienced test and study related worry and emotionality would also score higher on the Tension Reduction subscale of the RFD; (b) Men were predicted to report tension reduction as a stronger reason to drink than women, and women were hypothesized to exhibit higher levels of test and study worry and emotionality than men; and (c) underclass students were predicted to drink for different reasons than upperclass students. Results of the present study indicated that (a) Students who experienced higher levels of Test Emotionality, Test Worry, and Study Worry were more likely to report drinking to reduce tension than for social camaraderie or for mood enhancement; (b) Men reported tension reduction and mood enhancement as statistically significantly more important reasons to drink than women, and although not statistically significant, women reported higher levels of test and study worry and emotionality than men; and (c) reasons for drinking did not differ by year in college.

Interpretation of the results from the present study is straightforward. Test Emotionality, Test Worry, and Study Worry were predictive of Tension Reduction as a student reported reason for drinking. This is an intuitive finding supported by previous research, as students who worry excessively about examination performance are likely to experience higher arousal levels in the classroom. Further, it stands to reason that students experiencing concerns about their examination performance and who are worrying about an examination while studying would identify tension reduction as one reason to engage in substance use.

To our knowledge, this is the first study to explicitly use test and study worry and emotionality as predictors of college students' reasons for drinking. However, other research has examined related dimensions. Tyssen, Vaglum, Aasland, Gronvold, and Ekeberg (1998) found that students enrolled in a medical program reported consuming alcohol to alleviate the tension caused by their training. Park, Armeli, and Tennen (2004) found that college students are more likely to consume alcohol on days that are identified as being more stressful. Bishop, Bauer, and Becker (1998) reported that in male students seeking help from the university counseling center, those who reported moderate to high test anxiety also indicated more pressure by peers to drink and more difficulty controlling drinking. King, Bernardy, and Hauner (2003) found that female problem drinkers also reported higher levels of stressful life events and trait anxiety than male drinkers. And among alcohol misusers, Dunne, Galatopoulos, and Schipperheijn (1993) found that women reported higher levels of anxiety than comparable male counterparts.

Thus, a number of previous studies have indicated that students who experience anxiety symptoms may also report tension reduction as a main reason why they consume alcohol. However, other research has indicated that engaging in social interaction with others is often the main reason that students engage in alcohol consumption (e.g., Cronin, 1997). Although the current results suggested that a group of students who experience anxiety and worry in the academic environment consume alcohol as a method of coping with tension, it is reasonable to assume that not all individuals consume alcohol for the same reason. Previous research has indicated that a majority of excessive college student alcohol use may be related to the social environment of college life (e.g., Cronin, 1997; Park, 2004). However, the identification of a subset of students who report test and study anxiety and endorse tension reduction as the primary reason for alcohol consumption suggests intervention strategies for reducing alcohol use on college campuses. Specifically, anxiety management training that targets test and study anxiety could conceivably have added benefits in reducing alcohol consumption among this subset of students.

Further, the findings in the present study are also consistent with research that has identified gender differences in both test anxiety and reasons for drinking. Novak et al. (2003) reported that men tended to report tension reduction as a reason for engaging in drinking behavior more often than women. Crocker, Schmitt, and Tang (1988) and Williams (1996) both found that women demonstrated higher levels of test worry and emotionality than male counterparts in academic environments. Thus, it is plausible that other aspects of the college experience are linked to the reasons why women engage in alcohol consumption.

Men scored higher than women on all three subscales of the RFD across all academic years. It is therefore plausible that women engage in substance use for additional reasons. Recent research by Park (2004) that identified the positive consequences of drinking as a stronger motivation to consume was particularly salient in female respondents. Indeed, women in the Park study more frequently reported that drinking allowed them to meet new friends, have fun and socialize, and express themselves than did their male counterparts. Thus, perhaps future research can explore the gender specific differences in the relationship between academic-related anxiety and reasons for consuming alcohol.

Limitations of the Present Study

A few limitations of the present study should be noted. First, the students examined here were enrolled at a small, liberal arts college in the southeast and therefore may not be typical of students at larger institutions or in other parts of the United States. Second, because sensitivity to anxiety can be conceptualized as a personality characteristic, it is possible that the differences found in the present study could at least be partially accounted for based on personality differences. Future research should examine test and study worry and emotionality in the larger context of personality so that these potential confounds can be somewhat controlled. Third, because this is the first study to use the Test and Study Attitudes Inventory, a relatively novel measure of test and study worry and emotionality, it is difficult to know whether we were measuring state anxiety or trait anxiety with the instrument. Although the test authors purport that the instrument measures long-standing predilections toward test and study worry and emotionality (i.e., trait anxiety), it is possible that those students who were actively experiencing a stressful testing situation answered differently than they would have if such a situation was not present. Perhaps future research should attempt to control for this variable in an effort to further distill the relationship between test and study worry and emotionality and students reasons for drinking. Finally, the current study did not actually employ a measure of alcohol consumption, and thus, it is impossible to determine actual substance use patterns in relation to student reported reasons for drinking.

Future Research Directions

Future research may examine alcohol consumption among students who report tension reduction as their primary motive for engaging in substance use and who also exhibit high scores on the TSAI. It is likely that alcohol consumption will increase for this group of students around times associated with higher test anxiety, such as midterm and final examinations. Programs that help students manage anxiety and decrease their symptoms may result in decreases in both alcohol consumption and tension reduction as a reason for student drinking. Helping students manage stressful situations rather than using alcohol as an escape mechanism will likely help reduce the chances of other anxiety-related problems later in life, as research has shown that college students who consume alcohol during college to combat life stressors or to reduce tension are more likely to continue to consume alcohol after graduation and to experience alcohol-related problems due to their consumption habits (Perkins, 1999). In contrast, Perkins also found that students who consumed alcohol for the explicit purpose of social camaraderie were less likely to continue consuming alcohol after graduation. Thus, identifying student motivations for drinking might also aid in identifying those students at risk for developing substance use disorders after completing their education. Indeed, continued research on the motivations for student drinking coupled with the development of intervention programs to target students who seek to reduce anxiety through alcohol consumption may result in long-term effects of reducing alcohol misuse.

CONCLUSION

Excessive alcohol consumption on the nation's college campuses has reached the level of a national health problem. Excessive collegiate alcohol consumption is attributed to a wide variety of factors including personality traits such as sensation seeking, cognitive factors such as outcome expectancies and reasons for drinking, and environmental factors such as social norms and stressors associated with collegiate life. Successful intervention programs will not only need to target problem drinkers or campus life in general, but must also identify subsets of individuals who consume alcohol for specific reasons. The current study identified a segment of the student population who endorse tension reduction associated with test anxiety as a primary reason for alcohol use. This initial study suggests the need for developing alcohol prevention programs based on specific student needs, as participants who experienced the highest levels of Test Emotionality, Test Worry, and Study Worry were most likely to report consuming alcohol for the purpose of tension reduction. These results support the hypothesis that a subset of students consume alcohol as a form of self-medication, related directly to stressors associated with college life. Further, results supported small, but noteworthy gender differences on all of subscales of the RFD and the TSAI. Women reported higher levels of test and study worry and emotionality than men, and men reported tension reduction and mood enhancement as reasons for drinking to a larger degree than women. These results are largely consistent with previous studies showing a relationship between stress and alcohol consumption among college students.

Author Note

Kevin M. Kieffer, Christopher Cronin, Debra L. Gawet, Department of Psychology, Saint Leo University, Saint Leo, Florida.

This research was supported by grant 1 RO3 AA11585-01 from the National Institute on Alcohol Abuse and Alcoholism to Christopher Cronin, Principle Investigator.

Correspondence concerning this article should be addressed to Kevin M. Kieffer, Department of Psychology, MC 2127, Saint Leo University, P.O. Box 6665, Saint Leo, Florida 33574, USA. E-mail: [e-mail]; Phone: [phone]; Fax: [phone].

References

Bergman, H. C., & Harris, M. (1985). Substance abuse among young chronic patients. Psychosocial Rehabilitation Journal, 8(3), 49-54.

Billingham, R. E., Parillo, A.V., & Gross, C. (1993). Reasons given by college students for drinking: A discriminant analysis investigation. International Journal of the Addictions, 28, 793-802.

Bishop, J. B., Bauer, K.W., & Becker, E. T. (1998). A survey of counseling needs of male and female college students. Journal of College Student Development, 39, 205-210.

Borella, R, Bargellinin, A., Rovesti, S., Pinelli, M., Vivoli, R., Solfrini, V., & Vivoli, G. (1999). Emotional stability, anxiety, and natural killer activity under examination stress. Psychoneuroendocrinology, 24, 613-627.

Bryant, A. L., & Zimmerman, M. A. (2002). Examining the effects of academic beliefs and behaviors on changes in substance use among urban adolescents. Journal of Educational Psychology, 94, 621-637.

Camatta, C. D., & Nagoshi, C. T. (1995). Stress, depression, irrational beliefs, and alcohol use and problems in a college student sample. Alcoholism: Clinical and Experimental Research, 19(1), 142-146.

Carey, K. B., & Carey, M. P. (1995). Reasons for drinking among psychiatric outpatients: Relationship to drinking patterns. Psychology of Addictive Behaviors, 9, 251-257.

Christie, K., Burke, J., Regier, D., Rae, D., Boyd, J., & Locke, B. (1988). Epidemiological evidence for early onset of mental disorders and higher risk of drug abuse in young adults. American Journal of Psychiatry, 145, 971-975.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.

Crocker, L., Schmitt, A., & Tang, L. (1988). Test anxiety and standardized achievement test performance in middle school years. Measurement and Evaluation in Counseling and Development, 21, 149-157.

Cronin, C. (1997). Reasons for drinking versus outcome expectancies in the prediction of college student drinking. Substance Use and Misuse, 32, 1287-1311.

Cutter, H. S. G., & O'Farrell, T. J. (1984). Relationship between reasons for drinking and customary drinking behavior. Journal of Studies on Alcohol, 45, 321-325.

Damphousse, K., & Kaplan, H. B. (1998). Intervening processes between adolescent drug use and psychological distress: An examination of the self-medication hypothesis. Social Behavior and Personality, 26, 15-130.

Degenhardt, L., & Hall, W. (2001). The relationship between tobacco use, substance-use disorders and mental health: Results from the National Survey of Mental Health and Well-being. Nicotine & Tobacco Research, 3, 225-234.

Dunne, F. J., Galatopoulos, C., & Schipperheijn, J. (1993). Gender differences in psychiatric morbidity among alcohol misusers. Comprehensive Psychiatry, 34, 95-101.

Fan, X. (1996). A SAS program for assessing multivariate normality. Educational and Psychological Measurement, 56(4), 668-674.

Haden, T. L., & Edmundson, E. W. (1991). Personal and social motivations as predictors of substance use among college students. Journal of Drug Education, 21, 303-312.

Hansell, S., & White, H. R. (1991). Adolescent drug use, psychological distress, and physical symptoms. Journal of Health and Social Behavior, 32, 288-301.

Harris, L. M., & Cumming, S. R. (2003). An examination of the relationship between anxiety and performance on prospective and retrospective memory tasks. Australian Journal of Psychology, 55, 51-55.

Johnston, L. D., O'Malley, P. M., & Bachman, J. G. (1989). Drug use, drinking, and smoking: National survey results from high school, college, and young adult populations 1975-1988 (DHHS Publication ADM 89-1638). Washington, D.C.: U.S. Government Printing Office.

Kalodner, C. R., Delucia, J. L., & Ursprung, A.W. (1989). An examination of the tension reduction hypothesis: The relationship between anxiety and alcohol in college students. Addictive Behaviors, 14, 649-654.

King, A. C., Bernardy, N. C., & Hauner, K. (2003). Stressful events, personality, and mood disturbance: Gender differences in alcoholics and problem drinkers. Addictive Behaviors, 28, 171-187.

Klein, H. (1992). Self-reported reasons for why college students drink. Journal of Alcohol & Drug Education, 37, 14-28.

Marshall, G. D., Agarwal, S. K., Lloyd, C., Cohen, L., Henniger, E. M., & Morris, G. J. (1998). Cytokine dysregulation associated with exam stress in healthy medical students. Brain, Behavior, and Immunity, 12, 297-307.

McIlroy, D., Bunting, B., & Adamson, G. (2000). An evaluation of the factor structure and predictive utility of a test anxiety scale with reference to student's past performance and personality indices. The British Journal of Educational Psychology, 70, 17-33.

Mehlum, L. (1999). Alcohol and stress in Norwegian United Nations peacekeepers. Military Medicine, 164, 720-724.

Mensch, B. S., & Kandel, D. B. (1988). Do job conditions influence use of drugs? Journal of Health & Social Behavior, 29, 169-184.

Miller, B. E., Miller, M. N., Verhegge, R., Linville, H. H., & Pumariega, A.J. (2002). Alcohol misuse among college athletes: Self-medication for psychiatric symptoms? Journal of Drug Education, 32, 41-52.

Novak, A., Burgess, E. S., Clark, M., Zvolensky, M. J., & Brown, R. A. (2003). Anxiety sensitivity, self-reported motives for alcohol and nicotine use, and level of consumption. Anxiety Disorders, 17, 165-180.

Park, C. L. (2004). Positive and negative consequences of alcohol consumption in college students. Addictive Behaviors, 29(2) 311-321.

Park, C. L., Armeli, S., & Tennen, H. (2004). The daily stress and coping process and alcohol use among college students. Journal of Studies on Alcohol, 65, 126-136.

Peleg-Popko, O. (2002). Children's test anxiety and family interaction patterns. Anxiety, Stress and Coping: An International Journal, 15, 45-49.

Perkins, H. W. (1999). Stress-related alcohol use leads to problem drinking. The Brown University Digest of Addiction Theory and Application, 18, 6.

Schuckit, M., & Hesselbrock, V. (1996). Alcohol dependence and anxiety disorders: What is the relationship? American Journal of Psychiatry, 153, 139-140.

Spielberger, C. D., & Lunsford, G. D. (2001). Test and study attitudes inventory. Redwood City, CA: Mind Garden.

Stanley, M. A., Now, D. M., Bourland, S. L., Beck, J. G., & Averill, P. M. (2001). Assessing older adults with generalized anxiety: A replication and extension. Behaviour Research & Therapy, 39, 221-235.

Sud, A., & Prabha, I. (1995). Test anxiety and academic performance: Efficacy of cognitive/relaxation therapies. Psychological Studies, 40, 179-186.

Swendsen, J. D., Tennen, H., Carney, M. A., Affleck, G., Willard, A., & Hromi, A. (2000). Mood and alcohol consumption: An experience sampling test of the self-medication hypothesis. Journal of Abnormal Psychology, 109, 198-204.

Test, M. A., Wallisch, L. S., Allness, D.J., & Ripp, K. (1989). Substance use in young adults with schizophrenic disorders. Schizophrenia Bulletin, 15, 465-476.

Thompson, B. (1990). MULTINOR: A FORTRAN program that assists in evaluating multivariate normality. Educational and Psychological Measurement, 50, 845-848.

Tyssen, R., Vaglum, R, Aasland, O. G. Gronvold, N. T., & Ekeberg, O. (1998). Use of alcohol to cope with tension, and its relation to gender, years in medical school and hazardous drinking: A study of two nation-wide Norwegian samples of medical students. Addiction, 93, 1341-1350.

Wadee, A. A., Kuschke, R. H., Kometz, S., & Berk, M. (2001). Personality factors, stress, and immunity. Stress and Health, 17, 25-40.

Williams, J. E. (1996). Gender-related worry and emotionality test anxiety for high-achieving students. Psychology in the Schools, 33, 159-162.

Wood, M. D., Nagoshi, C. T., & Dennis, D. A. (1992). Alcohol norms and expectations as predictors of alcohol use and problems in a college student sample. Journal of Drug and Alcohol Abuse, 18, 461-476, 1992.

Yokoyama, K., Nishikitani, M., & Araki, S. (1999). Reasons for drinking in relation to problem drinking behavior in a sample of Japanese high school students. International Journal of Behavioral Medicine, 6, 135-149.

Kevin M. Kieffer, Christopher Cronin, Debra L. Gawet, Saint Leo University

Table 1
Canonical Solution for Test Worry and Emotionality
Predicting Students' Reasons for Drinking for Function 1

                                    Function 1

Variable                 Coef.   [r.sub.s]   [r.sup.2.sub.s]

Predictor Variables
    Test Worry           -.988     .546          29.81%
    Test Emotionality    1.288     .770          59.29%
    Study Worry           .957     .823          67.73%
    Study Emotionality   -.352     .681          46.38%
Criterion Variables
    Social Camaraderie    .069     .645          41.60%
    Mood Enhancement     -.161     .721          51.98%
    Tension Reduction    1.076     .996          99.20%
[R.sup.2.sub.c.]                                 10.30%

Note: Coef. = standardized canonical function coefficient;
[r.sub.s] = structure coefficient; [r.sup.2.sub.s] = squared
structure coefficient; [R.sup.2.sub.c] = squared canonical
correlation for Function 1.

Table 2
Means and Standard Deviations of RFD and TSAI Subscales
by Gender and Year in College (n = 365)

                           Men             Women

Subscale               M      (SD)      M      (SD)

Test Worry
    Freshman         15.53   (5.72)   16.29   (6.27)
    Sophomore        14.28   (5.38)   16.27   (5.91)
    Junior           14.61   (3.88)   16.45   (6.23)
    Senior           13.88   (5.30)   15.38   (6.29)
    Total            14.64   (4.95)   16.17   (6.12)
Test Emotionality
    Freshman         15.93   (5.59)   16.64   (6.64)
    Sophomore        13.83   (5.71)   16.58   (6.13)
    Junior           15.94   (6.21)   16.26   (6.00)
    Senior           13.25   (4.06)   16.58   (6.77)
    Total            14.93   (5.64)   16.54   (6.35)
Study Worry
    Freshman         17.33   (4.88)   17.73   (4.93)
    Sophomore        14.11   (5.59)   16.82   (4.92)
    Junior           14.22   (3.96)   17.19   (6.20)
    Senior           15.12   (4.88)   15.50   (5.26)
    Total            15.10   (4.92)   17.04   (5.25)
Study Emotionality
    Freshman         15.67   (5.52)   15.69   (5.99)
    Sophomore        13.22   (6.01)   15.22   (5.62)
    Junior           14.11   (5.03)   15.61   (6.17)
    Senior           12.50   (4.20)   14.42   (5.67)
    Total            14.02   (5.36)   15.35   (5.84)
Social Camaraderie
    Freshman          6.13   (4.39)    5.02   (5.01)
    Sophomore         5.94   (4.76)    4.44   (4.63)
    Junior            7.67   (5.75)    6.58   (3.95)
    Senior            6.00   (4.63)    4.23   (3.87)
    Total             6.52   (4.91)    5.03   (4.59)
Mood Enhancement
    Freshman          4.67   (4.33)    2.27   (3.29)
    Sophomore         4.38   (4.95)    2.26   (3.33)
    Junior            5.44   (5.17)    3.16   (3.30)
    Senior            3.75   (2.96)    2.15   (3.32)
    Total             4.69   (4.58)    2.42   (3.30)
Tension Reduction
    Freshman          4.33   (3.97)    2.03   (2.92)
    Sophomore         3.78   (4.28)    2.40   (3.20)
    Junior            4.72   (6.75)    3.00   (3.02)
    Senior            3.50   (2.62)    2.65   (2.88)
    Total             4.17   (4.86)    2.41   (3.01)

                          Total

Subscale               M      (SD)    Cohen's d (a)

Test Worry
    Freshman         16.15   (6.14)       -.121
    Sophomore        15.69   (5.79)       -.353
    Junior           15.78   (5.51)       -.364
    Senior           15.03   (6.03)       -.259
    Total            15.77   (5.86)       -.276
Test Emotionality
    Freshman         16.51   (6.42)       -.116
    Sophomore        15.79   (6.10)       -.465
    Junior           16.14   (6.01)       -.052
    Senior           15.79   (6.34)       -.615
    Total            16.12   (6.20)       -.269
Study Worry
    Freshman         17.66   (4.89)       -.082
    Sophomore        16.05   (5.23)       -.516
    Junior           16.10   (5.63)       -.585
    Senior           15.41   (5.10)       -.075
    Total            16.53   (5.22)       -.382
Study Emotionality
    Freshman         15.68   (5.87)       -.003
    Sophomore        14.65   (5.76)       -.344
    Junior           15.06   (5.77)       -.268
    Senior           13.97   (5.37)       -.389
    Total            15.00   (5.74)       -.238
Social Camaraderie
    Freshman          5.20   (4.89)        .236
    Sophomore         4.87   (4.68)        .319
    Junior            6.98   (4.67)        .225
    Senior            4.65   (4.06)        .416
    Total             5.42   (4.71)        .314
Mood Enhancement
    Freshman          2.72   (3.61)        .629
    Sophomore         2.87   (3.94)        .512
    Junior            4.00   (4.18)        .538
    Senior            2.52   (3.27)        .509
    Total             3.01   (3.80)        .576
Tension Reduction
    Freshman          2.47   (3.24)        .668
    Sophomore         2.79   (3.57)        .369
    Junior            3.63   (4.75)        .352
    Senior            2.85   (2.80)        .309
    Total             2.81   (3.66)        .447

Note: N = 120 for Freshman, 102 for Sophomores, 81 for juniors and
63 for Seniors; N = 118 for Men, 247 for Women.

(a) Cohen's d is calculated as ([Mean.sub.men]-[Mean.sub.women])/
(([SD.sub.men] + [SD.sub.women])/2); Sign of the standardized mean
difference is arbitrary and is governed only by which mean is placed
first in the equation (i.e., men and women could be transposed to
generate a Cohen's d with the opposite sign).