Expectancies specific to condom use mediate the alcohol and sexual risk relationship.

Young adults are one of the fastest growing groups of new cases of HIV\AIDS and other STDs in the United States (Wasserheit, 1997). Yet they continue to engage in high-risk sexual activities, such as failing to use a condom during sexual intercourse (Diclemente, 2000). Since AIDS and other STDs are primarily transmitted by sexual activity, research has examined correlates of this activity to decrease sexual risk-taking. Drinking is one correlate of sexual activity that has received much attention.

A well-established relationship between drinking and sexual activity exists in both popular culture and research. Alcohol use is related to earlier and increased incidences of sexual behavior, greater numbers of sexual partners, and greater intentions to engage in sex (Crowe & George, 1989; Epstein, Dusenbury, Botvin, & Diaz, 1994; Temple & Leigh, 1992). In addition, drinking may contribute to risky sexual behavior. Public AIDS prevention campaigns have directly targeted this proposed relationship. Alcohol use may increase the likelihood of HIV and other STD infections by decreasing the likelihood of using a condom during sex. This proposed effect of drinking is called the transmission hypothesis (Dingle & Oei, 1997).

Evidence for the transmission hypothesis was found in several studies that revealed associations between alcohol use and sexual risk-taking (e.g., Graves, 1995; Leigh & Stall 1992; McEwan, McCallum, Bhopal, & Madhok, 1992; Meilman, 1993). However, a similar association was not revealed in all studies (e.g., Gold & Skinner, 1993; Leigh, Temple, & Trocki, 1994; Weatherburn, Davies, Hickson, & Hunt, 1993). Two recent reviews of studies examining the proposed alcohol and risky sex link (Dingle & Oei, 1997; Halpern-Felsher, Millstein, & Ellen, 1996) confirmed that while there is partial support for the transmission hypothesis in the literature, it is not universal. Dingle & Oei described the findings as equivocal. Given this conflicting and equivocal evidence, researchers have focused on potential mediators that might help explain the relationship between alcohol use and risky sex. Understanding mechanisms through which alcohol adversely affects safer-sex intentions and behaviors is important for reducing risk of contracting HIV and other STDs.

Expectancy Theory and Safer Sex Models

Following the Theory of Reasoned Action (Ajzen, 1971) and the Theory of Planned Behavior (Ajzen, 1991), sex researchers have predicted condom intention as a way of determining sexual risk (e.g., Corby, Schneider-Jamner, & Wolitski, 1996; Morrison, Gillmore, & Baker, 1995; Sneed & Morisky, 1998). These theories propose that behavior is a direct result of behavioral intentions. Consistent with the theories, condom intention has been found to be a good predictor of prospective safer sex behaviors (Sheeran & Orbell, 1998). Both theories incorporate the concept of anticipated outcomes of risk-reducing behavior under their concept of attitudes, which in turn predict intentions. Other models also utilize expected outcomes. Social cognitive theory (Bandura, 1986) proposes that health-protective behavior results from a process of cognitive appraisal integrating knowledge, outcome expectancies associated with adopting risk-reduction behavior, and social influences.

The focus on outcome expectancies in models of sexual risk behavior seems particularly pertinent when examining the drinking and risky sex link given the growing interest in cognitive processes and interventions, as well as the large alcohol expectancy literature (e.g., Goldman, 1994). Tolman (1932) first introduced the expectancy concept in learning theory. Later work by Bolles (1979) and Kirsch (1985) proposed that expectancies were cognitive variables that represented information about the relationship of a response to a stimulus outcome or between a stimulus event and a subsequent response. The activation of expectancies can directly initiate a behavioral sequence previously associated with the recognized stimulus (Goldman, Del Boca, & Darkes, 1999). Expectancy theory proposes that expectancies influence almost all behavior. Thus, they can be applied to drinking and drinking-induced behavior. In alcohol research, expectancies are beliefs about the anticipated (expected) effects alcohol will have in a future situation (Roehrich & Kinder, 1991). Most of the research on expectancies has attempted to show how positive general expectancies predict drinking patterns (e.g., Marlatt & Gordon, 1985). Less attention, however, has been paid to the important possibility that alcohol expectancies may influence post-drinking behavior.

Expectancies as a Mediator of Alcohol's Relationship to Subsequent Behavior

Baron and Kenny (1986) claimed that internal cognitive processes mediate the effects of external stimuli (such as alcohol) on behavior. Mediators help explain the internal psychological processes of how or why effects occur. Expectancies are one such internal cognitive process. The possibility that post-drinking behaviors could be driven by alcohol expectancies independent of pharmacological effects was suggested by a series of placebo-type studies (Marlatt & Rohsenow, 1980).

Theoretically, individual differences in expected impairment of behavior should predict the degree of impairment under alcohol. Those who expect impairment should be more impaired when they drink than those who do not expect impairment (Vogel-Sprott & Fillmore, 1999). In a series of studies, persons who expected the most impairment had larger performance drops on cognitive and psychomotor tasks, both when they received alcohol and when they received a placebo (Fillmore, Carscadden, & Vogel-Sprott, 1998; Fillmore & Vogel-Sprott, 1995, 1996). These relationships between expectancies and alcohol-related behavior suggest that expectancies mediate post-drinking behavior. To actually infer causality, however, it must be shown that manipulating expectancies about alcohol's effects alters alcohol-related behavior. Studies using placebo found that manipulating expectancies altered performance (Fillmore, Mulvihill, & Vogel-Sprott, 1994; Fillmore & Vogel-Sprott, 1992). The altered performance could only be explained by the new expectancies. In concluding their chapter on expectancy and the behavioral effects of alcohol use, Vogel-Sprott and Fillmore (1999) stated that interpersonal and intersitutational differences in expectancies help account for differences in behavioral responses to alcohol that could not otherwise be explained by pharmacological action. Thus, empirical evidence supports the hypothesis that expectancies mediate alcohol-related behavior.

A Proposed Mediator--Alcohol Expectancies for Condom Use

Expectancies specific to alcohol's effects on risky sexual behaviors may help explain any observed relationships between drinking and risky sex behaviors, as well as help explain previous inconsistent findings. Dermen and Cooper (1994) developed a scale of sexual risk-taking expectancies that included several aspects of risk-taking such as disinhibited sexual behavior and failure to practice safer sex. However, people may have expectancies solely around alcohol's effect on condom use: in particular, about desires to use a condom, as well as abilities to negotiate condom use and actually use a condom after drinking. These are different from general sexual risk-taking expectancies because they specifically target condom use, a marker of sexual risk for those who are sexually active. These specific expectancies, which we propose to study for the first time here, are better suited for explaining the alcohol and condom use link because they directly target the risk behavior of interest. In sexual situations involving drinking, these expectations may influence condom use skills.

The Present Study

We tested a model in which alcohol expectancies for condom use were hypothesized to mediate the relationship between alcohol use and sexual risk intentions in sexually active college students. We hypothesized that these expectancies would predict sexual risk (defined as intention to use a condom in future sex events involving drinking), as well as mediate the relationship between alcohol use and the sexual risk. We controlled for past condom use when drinking. Past condom use is often the best predictor of future condom use (Vasilias, 1999). We hypothesized that the expectancy-mediated relationship between alcohol use and sexual risk intentions would emerge even with control for the part of condom use intention explained by past behavior.

METHOD

Participants

Recruitment and procedure. Male and female heterosexual college students (N = 563; Females = 336, Males = 227) from two universities, one an urban private university and the other a rural state university, participated in this study. They were recruited through their psychology classes or through flyers to campus organizations. The students from the psychology courses received course credit for their participation. The students recruited through the flyers received a nominal stipend ($5) for participation. Participants came to a classroom on their own campus designated for the study and consented to participate in the study. They then completed a confidential questionnaire about their attitudes, beliefs, and behavioral habits with respect to drinking and sex. While the study hypothesized relationships for sexually active persons (those who had engaged in vaginal intercourse in the previous 6-month period), the nonsexually active participants were encouraged to answer attitudinal, expectancy, and intention questions by imagining how alcohol would affect them in sexual situations.

Demographics. Sixty-two percent (N = 347) of the participants engaged in vaginal intercourse in the 6-month period prior to participation in the study (Females: 57%, N = 192; Males: 68%, N = 155). To capture the dynamics of the alcohol use and sexual risk relationship, only these sexually active participants were included in the analyses. There were no differences on age, ethnicity or site of study between the sexually active men and women. All sexually active participants ranged in age from 17 to 26 (mean = 20.3 SD = 3.50). Fifty-eight percent were Caucasian, 16% Hispanic, 12% Asian, 6% African-American, and 8% other. Furthermore, there were no age or ethnic differences between sexually active and nonsexually active participants on the variables of interest in this study.

Measures

Alcohol use. Participants completed a quantity and frequency index for the alcohol use within the past 6 months. They were asked, "How many days per week on average do you drink?" and "How many drinks on average do you have when you drink?" They were told that a drink included a 1-ounce shot of liquor alone or in a mixed drink, one beer, or a 4-ounce glass of wine. The two items were multiplied to produce the quantity x frequency Alcohol Use index used in the analyses. Similar measures were used in previous studies and shown to be valid indices of alcohol consumption (Earleywine & Martin, 1993).

Alcohol expectancies for condom use. Participants responded to three items developed for this questionnaire that asked how they expected alcohol to affect negotiating condom use and using a condom, rating each on a Likert scale (1 = Not at all like me/Never true and 7 = Totally like me/Always true). These items were: (a) After a few drinks it is more difficult for me to talk to a potential sexual partner about using a condom; (b) Drinking makes me forget about using condoms; and (c) When I drink, it is hard to use (or have my partner use) a condom. Higher scores represent expectancies that alcohol more negatively affects one's ability to negotiate and use condoms. These items were averaged to form the overall alcohol expectancies for condom use scale. This scale had adequate reliability ([alpha] = .84, 3 items, N = 328).

Condom use. Participants responded to an item that asked how often in the past 6 months they had used a condom when they were drinking prior to sex. They responded on a Likert scale (1 = Never to 5 = Always).

Intention to use a condom. Participants responded to an item which asked how likely it was that they would use a condom in the future if they had been drinking. They were instructed to respond even if they did not intend to have sex in the near future, as well as if they did not intend to drink prior to future sexual events. They responded on a Likert scale (1 = Will never use a condom to 5 = Will use a condom always).

RESULTS

We proposed an expectancy-mediated model in which alcohol expectancies for condom use mediate the relationship between alcohol use and behavioral intention. Table I contains the means and standard deviations for sexually active males and females on the drinking, expectancies, and sexual risk variables. There were no gender differences on age, past condom use when drinking, and intention to use a condom if drinking. However, men had more sexual partners in the past 6 months (F(1,343) = 18.70, p < .001) and more lifetime sexual partners (F(1,337) = 14.59, p < .001) than women. They also drank more (F(1,342) = 39.75, p < .001) and expected drinking to more negatively impact condom use (F(1,331) = 27.28, p < .001).

Correlations

The differences between men and women on the alcohol use and expectancies variables suggest that relationships between them in the mediational analyses may also be different for men and women. Table 2 provides correlations among the variables for both the sexually active men and the sexually active women. A number of the correlations are statistically significant. For mediation to be possible, there must be a significant relationship between alcohol use and intention. Drinking must also be related to expectancies for condom use, while these expectancies must be related to intention for condom use (Baron & Kenny, 1986). This is the case with the men. Drinking is related to condom use expectancies and future intention to use a condom if drinking, while expectancies are related to condom intention. There is no direct relationship, however, between alcohol use and intention in females, nor is there a relationship between alcohol use and expectancies for condom use. This implies different gender pathways for the link between alcohol use and sexual risk intentions in this sample.

Regression Analyses

Table 3 contains the results from estimated hierarchical regression equations that test the expectancy-mediation model in men. Because there was no direct relationship between alcohol use and condom intention in women, there was nothing to mediate. Therefore, there was no need to conduct these analyses on the women's data. The outcome variable in the regression equations was intention to use a condom when drinking in the future. In the analyses, past condom use when drinking served as a covariate. Thus, any revealed relationship (direct or indirect) between alcohol use and intention would be over and above the variance accounted for by past behavior. Controlling for past behavior decreases the likelihood that any alcohol-risky sex links that emerge can be explained by past behavior, condom use, and other factors that influenced past condom use when drinking.

In the first reported regression equation, past condom use when drinking, the covariate, was entered on Step 1 and alcohol use was entered on Step 2. In this sample of sexually active men, alcohol use had a direct effect on condom intention in future drinking situations over and above the effect of past behavior. Those who drank more had lower intention to use a condom in future drinking situations. This equation explained 21% of the variance in intention for men.

In the second regression equation, past condom use when drinking was entered on Step 1, alcohol expectancies for condom use were entered on Step 2, and alcohol use was entered on Step 3. With condom use expectancies now in the equation, the previously revealed direct effect for alcohol use on intention disappeared. The expectancies for condom use, however, predicted intention even when covarying out past condom use. Expectations that drinking adversely influences condom use predicted lower intention to use a condom. The three predictors (alcohol use, alcohol expectancies for condom use, and past condom use when drinking) accounted for 30% of the variance in intention.

Thus, these results are consistent with Baron and Kenny's (1986) description of mediator operation. The difference in [beta] (path weights) for alcohol use in the two regression equations was significant (t = -2.9, p < .01), implying that the alcohol expectancies accounted for a significant part of the drinking and risky sex link. Using the correlation between alcohol use and expectancies along with the second regression analysis, a significant indirect effect for alcohol use on intention through alcohol expectancies for condom use was found (t = -2.8, p < .01).

DISCUSSION

Alcohol expectancies specific to condom use are a new construct that may be important for understanding the link between alcohol use and condom use. These specific expectancies mediated the relationship between alcohol use and sexual risk intentions in sexually active college men. The more a young man drank, the more he expected alcohol to impair his condom use skills. These expectancies, in turn, predicted lower intention to use a condom in sexual experiences involving drinking. In women, however, neither alcohol use nor condom use expectancy was related to condom intention. The results in the men were revealed even while controlling for past condom use in sex events involving drinking. This control is significant since past condom use is the most powerful predictor of both condom intentions and future condom use (Brook et al., 2000; Vasilias, 1999). Yet we cannot manipulate or intervene at past behavior. In men, however, alcohol use indirectly predicted condom intention in future drinking situations through expectancies over and above the portion of intention accounted for by past behavior.

While it would be preferable to study actual condom use prospectively, condom intentions are important because condom use is the direct result of these intentions (Sheeran, & Orbell, 1998; Sneed & Morisky, 1998). Several experiments have found that drinking decreases intention to engage in safer sex behaviors (Gordon, Carey, & Carey, 1997; MacDonald, Fong, Zanna, & Martineau, 2000; MacDonald, Zanna, & Fong, 1996). Models predicting condom use intentions have found that several variables, including social norms, attitudes, and perceived behavioral control, predict condom intention (Corby et al., 1996; Morrison et al., 1995; Sneed & Morisky, 1998). Past condom use also predicts intentions (Maticka-Tyndale & Herold, 1999). Moreover, some of the strongest predictors of intention are situational factors, such as partner type, availability of condoms, alcohol use, and situational expectancies or cognitions (Herold, Maticka-Tyndale, & Mewhinney, 1998; Maticka-Tyndale & Herold, 1999; Raj, 1996). These variables in various combinations predict between 10% and 65% of the variance in intention. Raj (1996) stated that models of safer sex must incorporate situational and social cognitive factors into their design. Our model did this, first by looking at the relationship between alcohol, a situational variable, and condom intention. Then, using expectancy theory (Goldman, 1994), we incorporated a new cognitive factor, expectancies specific to condom use, into the model as a potential explanation for how alcohol use affects sexual risk intentions.

How Expectancies Work

One potential limitation of this expectancy-mediated model is its inability to capture the reciprocal effects of drinking and expectancy. That is, while drinking may predict and determine expectancies as presented in the model, it is possible that expectancies also predict future alcohol consumption. Expectancies about the general effects of alcohol, such as a positive impact on socialization, have been found to predict alcohol use (Brown, 1985; Martin & Hoffman, 1993). People may drink because they think it enhances their ability to relate to others, gives them social courage, and makes them more interesting and even more sexually appealing. However, it is reasonable to assume this is not the case here. The expectancies in this study are specific to alcohol's impact on condom use. Stacy (1997) argues for using specific expectancies in behavioral models. It is highly unlikely that people drink because of these specific expectancies. That is, they are unlikely to drink in order to decrease their ability to negotiate condom use or use a condom.

The present study suggests that past alcohol consumption impacts expectancies for condom use. Learning theory provides an explanation for how this occurs. Persons store relatively specific memories of prior experience and learning about alcohol's effect on their behavior (Goldman, 1994; Vogel-Sprott & Fillmore, 1999). Thus, past drinking experiences help form alcohol expectancies. They can also be formed, however, through observation and vicarious learning (Goldman, 1994). On subsequent drinking occasions, the activation of this stored specific information provides a template for the emission of similar behavior. This is how expectancies impact post-drinking behavior. While we did not experimentally show mediation of alcohol and sexual risk in real sex events involving drinking, we believe that the revealed expectancy-mediated effects will be present in these situations.

In theory, the presence of alcohol causes the activation of expectancies in memory. Once activated, they influence thoughts, interpretations, behavioral intentions, decisions, and behaviors (Stacy, 1997). Alcohol may serve as a cue, making certain behaviors more accessible and likely. Stacy proposed that deliberate considerations do not govern all the processes underlying this mechanism. Expectancy theorists have suggested recently that expectancies exert ongoing control of behavior in real-life environments, and that expectancies are not exclusively either conscious or automatic (Goldman, Del Boca, & Darkes, 1999). Rather, they are spontaneously activated and then through both conscious and automatic means impact thoughts and behavioral decisions. Thus, it is expected that in a sex event involving drinking, expectancies specific to condom use could be activated and impact actual condom use.

Gender Differences

There were different gender pathways in the alcohol and sexual risk relationship in this sample. In the women, no relationship between alcohol use and sexual risk intentions existed. There was also no relationship between expectancies and intention for the women. McKay (1999) also found that expectancies failed to predict sexual risk in women. Gender differences in the relationships between alcohol and various sexual risk behaviors have been revealed in previous research (e.g., Anderson & Mathieu, 1996; MacDonald et al., 1996; Murphy, Rotheram-Borus, & Reid, 1998). These gender differences, however, do not follow any consistent direction. Some studies show increased risk for women, others show decreased risk for women, and still others show no gender differences in risk after drinking.

Bell, O'Neal, Feng, and Schoenrock (1999) tested two models for gender differences in sexual risk-taking: difference in magnitude versus difference in patterns. They found support for the different patterns hypothesis suggesting that women may engage or not engage in sexual risk-taking for different reasons than men. One potential explanation for these different patterns is birth control. Unwanted pregnancy appears to be a far more salient consequence of risky sexual behavior for women than for men. Alcohol use may not impact their sexual risk taking if they are using condoms primarily for pregnancy prevention. If women are not using condoms primarily for birth control, they may be less inclined to use condoms.

Alternatively, the difference in numbers of partners between men and women may suggest sexual context differences. Women may not engage in casual sex as often as men. Oliver and Hyde (1993) found that women endorse less positive attitudes toward casual sex than do men. Condom use tends to occur earlier rather than later in relationship development. Thus, fewer women may intend to be in situations where they are likely to drink, have sex, and need to use a condom. Perhaps the failure to find a link between alcohol and risky sex intentions in women results from women thinking that if they drink and have sex it will be in the context of a well-established relationship where condom use is not an issue. Either they have decided to stop using condoms or maybe they never used condoms in the first place. Furthermore, O'Hare (1999) found a gender X context interaction for risky sexual beliefs and behaviors. Nonetheless, the revealed gender differences in the alcohol use and risky-sex link here and elsewhere require further study. Finally, future studies must be cognizant of gender differences in sexual risk intentions and behaviors.

Implications and Expectancy Challenges

The mediational role of expectancies for condom use together with the already established disinhibitory relationship between drinking and sexual activity increases concern. Young adults who drink more heavily are also more likely to have higher numbers of sexual partners and, therefore, more occasions for potential HIV and other STD transmission (Biglan et al., 1990). Heavier drinkers also appear to expect drinking to decrease their condom use skills. Future interventions could target irresponsible drinking, as well as expectancies for condom use that are associated in men with decreased condom intention.

Challenges to alcohol expectancies have reduced both drinking and post-drinking behavioral impairment (Darkes & Goldman, 1993). Darkes and Goldman (1998) found that challenging expectancies for alcohol's impact on sociability and arousal reduced alcohol consumption. Furthermore, strong challenges to expectancies of impairment on cognitive and psychomotor tasks lead to less impairment (Fillmore & Vogel-Sprott, 1995). These findings lend support to the theoretical notion of expectancy mediation. They also suggest challenges to expectancies of condom use impairment might be successful in reducing the alcohol use and risky sex link. Research and intervention programs that intensely challenge specific expectancies that alcohol impairs one's ability to negotiate use and to use a condom might build efficacy for condom use skills in future drinking situations. These challenges could be particularly helpful with persons who use being drunk as an excuse for risky sexual behavior. Finally, interventions that help people negotiate condom use and use condoms when drinking could lower expectancies of impairment and, thus, seemingly decrease sexual risk-taking.

Limitations

The measurement of alcohol use is a limitation of the study. We measured an index of participants' general drinking patterns, but did not directly assess drinking in sexual situations. We do not know if the drinking and sexual behaviors occur temporally together. Perhaps some persons drink heavily in nonsexual situations only and don't drink at all in sexual situations. However, a general drinking index may actually reflect drinking patterns in sexual situations. Many studies (Dingle & Oei, 1997) of the alcohol use and risky-sex link use correlational data similar to that employed in this study. It would be particularly helpful for future studies to assess drinking in real sex events. Nonetheless, experimental evidence exists that drinking decreases motivation and intention to use a condom (Gordon et al., 1997; MacDonald et al., 2000; MacDonald et al., 1996). The present findings can be interpreted in this light.

Due to the sample of college students, the generalizibility of these results is in question. Age may moderate the examined relationships. Reviews found that there tends to be a relationship between drinking and sexual risk in adolescents and college-aged persons, but not in adults (Dingle & Oei, 1997; Halpern-Felsher, Millstein, & Ellen, 1996; LaBrie, 2000). Older persons may be more experienced drinkers. Expectancies did not mediate the relationship between alcohol use and subsequent behavior in experienced drinkers (Vogel-Sprott & Fillmore, 1999). Nonetheless, the growing rate of HIV transmission among young adults coupled with high rates of drinking in college students (Nicklin, 1999; Reisberg, 1998) makes this sample relevant.

Conclusion

Finally, our findings provide partial support for the transmission hypothesis. Alcohol use decreases the likelihood of condom use in college-aged men and, thus, increases the likelihood of HIV and other STD infections. We also introduced a potential explanation for this increased risk among male drinkers--expectancy mediation. Alcohol expectancies for condom use help explain the relationship between alcohol consumption and sexual risk. These expectancies are new and specific to an important safer-sex behavior: condom use. As a revealed mediator of the alcohol use and risky-sex link, they may be important in uncovering the role drinking plays in sexual risk-taking and in reducing these behaviors. Further research could aid in developing a rigorous scale for these expectancies, as well as testing their durability, how they are activated, and whether differing levels of activation relate to different effects. This study suggests that expectancies that alcohol impairs condom use skills are important and might lead to increased sexual risk-taking when drinking. Ironically, messages that build and promote these expectancies may actually be counterproductive, helping to increase risky sex in drinkers. Public health officials might be better served by educating young people that they can use condoms when they drink.

Table 1. Drinking and Sexual Risk Variables for Sexually
Active Participants

                                                Men

                                           Mean        SD

Age                                      20.6          2.98
Sexual Partners (past 6 mos.)             2.47 ***     2.71
Sexual Partners (lifetime)                8.02 ***    11.91
Alcohol use                              15.51 ***    16.97
Condom use when drinking                  3.39         1.50
Alcohol expectancies for condom use       3.10 ***     1.72
Intention if drinking                     5.53         1.71

                                             Women

                                         Mean      SD

Age                                      20.1      3.45
Sexual Partners (past 6 mos.)             1.50     1.39
Sexual Partners (lifetime)                4.34     5.08
Alcohol use                               6.20    10.24
Condom use when drinking                  3.50     1.57
Alcohol expectancies for condom use       2.63     1.29
Intention if drinking                     5.77     1.76

*** p < .001.
Table 2. Correlations Among Variables For Sexually Active Men and Women

                                                       Condom use
                                       Alcohol use    when drinking

Alcohol use                                 --            -.13
Condom use when drinking                   .18 *            --
Alcohol expectancies for condom use        .14            -.15 *
Intention if drinking                      .11             .31 ***

                                                      Intention if
                                           AECU          drinking

Alcohol use                                .27 ***        -.23 **
Condom use when drinking                  -.04             .37 ***
Alcohol expectancies for condom use         --            -.34 ***
Intention if drinking                     -.24 **           --

Note. Correlations for the sexually active men are contained on the
upper half of the table, while correlations for the sexually active
women are contained on the lower half of the table.

* p < .05. ** p < .01. *** p <.001.

REFERENCES

Ajzen, I. (1971). Attitudinal vs. normative messages: An investigation of the differential effects of persuasive communications on behavior. Sociometry, 34, 263-280.

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179-211.

Anderson, P. B., & Mathieu, D. A. (1996). College student's high-risk sexual behavior following alcohol consumption. Journal of Sex and Marital Therapy, 22, 259-265:

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall.

Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182.

Bell, N. J., O'Neal, K. K., Feng, D., & Schoenrock, C. J. (1999). Gender and sexual risk. Sex Roles, 41, 313-332.

Biglan, A., Metzler, C. W., Wilt, R., Ary, D., Noell, J., Ochs, L., French, C., & Hood, D. (1990). Social and behavioral factors associated with high-risk sexual behavior among adolescents. Journal of Behavioral Medicine, 13, 245-261.

Bolles, R. C. (1979). Learning theory (2nd ed.). New York: Holt, Reinhart, & Winston.

Brook. D. W., Brook, J. S., Rosenberg, G., Whiteman, M., Masci, J. R., Roberto, J., & de Cotalogne, J. (2000). Longitudinal pathways to condom use: A psychosocial study of IDUs. Journal of Addictive Diseases, 19, 55-69.

Brown, S. A. (1985). Expectancies versus background in the prediction of college drinking patterns. Journal of Counseling and Clinical Psychology, 53, 123-130.

Corby, N., Schneider-Jamner, M., & Wolitski, R. J. (1996). Using the theory of planned behavior to predict intention to use condoms among male and female injecting drug users. Journal of Applied Social Psychology; 26, 52-75.

Crowe, L. C., & George, W. H. (1989). Alcohol and human sexuality: Review and integration. Psychological Bulletin, 105, 374-386.

Darkes, J., & Goldman, M. S. (1993). Expectancy challenge and drinking reduction: Experimental evidence for a mediational process. Journal of Consulting and Clinical Psychology, 61, 344-353.

Darkes, J., & Goldman, M. S. (1998). Expectancy challenge and drinking reduction: Process and structure in the alcohol expectancy network. Experimental and Clinical Psychopharmacology, 51, 64-76.

Dermen, K. H., & Cooper, M. L. (1994). Sex-related alcohol expectancies among adolescents: I. Scale development. Psychology of Addictive Behaviors, 8, 152-160.

Diclemente, R. J. (2000). Looking forward: Future direction for HIV prevention research. In J. Peterson & R. J. Diclemente (Eds.), Handbook of HIV prevention (pp. 311-324). New York: Plenum.

Dingle, G. A., & Oei, T. P. S. (1997). Is alcohol a cofactor of HIV/AIDS? Evidence from immunological and behavioral studies. Psychological Bulletin, 122, 56-77.

Earleywine, M., & Martin, C. (1993). Anticipated stimulant and sedative effects of ethanol vary with limb of the blood alcohol curve and dosage. Alcoholism: Clinical and Experimental Research, 51, 135-139.

Epstein, J., Dusenbury, L., Botvin, G., & Diaz, T. (1994). Determinants of intentions of junior high school students to become sexually active and use condoms: Implications for reduction and prevention of AIDS risk. Psychological Reports, 75, 1043-1053.

Fillmore, M. T., Carscadden, J. L., & Vogel-Sprott, M. (1998). Alcohol, cognitive impairment and expectancies. Journal of Studies on Alcohol, 59, 174-179.

Fillmore, M. T., Mulvihill, L. E., & Vogel-Sprott, M. (1994). Psychomotor performance under alcohol and caffeine: Pharmacological and expectancy effects. Experimental and Clinical Psychopharmacology, 3, 33-38.

Fillmore, M. T., & Vogel-Sprott, M. (1992). Expected effect of caffeine on motor performance predicts type of response to placebo. Psychopharmacology, 122, 175-181.

Fillmore, M. T., & Vogel-Sprott, M. (1995). Expectancies about alcohol-induced motor impairment predict individual differences in responses to alcohol and placebo. Journal of Studies on Alcohol, 56, 90-98.

Fillmore, M. T., & Vogel-Sprott, M. (1996). Evidence that expectancies mediate behavioral impairment under alcohol. Journal of Studies on Alcohol, 57, 598-603.

Gold, R. S., & Skinner, M. J. (1993). Desire for unprotected intercourse preceding its occurrence: The case of young gay men with an anonymous partner, International Journal of STD and AIDS, 4, 326-329.

Goldman, M. S. (1994). The alcohol expectancy concept: Applications to assessment, prevention, and treatment of alcohol abuse. Applied and Preventive Psychology, 3, 131-144.

Goldman, M. S., Del Boca, F. K., & Darkes, J. (1999). Alcohol expectancy theory: The application of cognitive neuroscience. In K. Leonard & H. T. Blane (Eds.), Psychological theories of drinking and alcoholism (pp. 203-246). New York: Guilford.

Gordon, C. M., Carey, M. E, & Carey, K. B. (1997). Effects of a drinking event on behavioral skills and condom attitudes in men: Implications for HIV risk from a controlled experiment. Health Psychology, 16, 490-495. Graves, K. L. (1995). Risky sexual behavior and alcohol use among young adults. American Journal of Health Promotion, 10, 27-36.

Halpern-Felsher, B. L., Millstein, S. G., & Ellen, J. M. (1996). Relationship of alcohol use and risky sexual behavior: A review and analysis of findings. Journal of Adolescent Health, 19, 331-336.

Herold, E. S., Maticka-Tyndale, E., & Mewhinney, D., (1998). Predicting intention to engage in casual sex. Journal of Social and Personal Relationships, 15, 502-516.

Kirsch, I. (1985). Response expectancy as a determinant of experience and behavior. American Psychologist, 102, 319-323.

LaBrie, J. W. (2000). The relationship of alcohol and sexual risk behaviors: A critical evaluation of event-level studies. Unpublished doctoral dissertation, University of Southern California.

Leigh, B. C., & Stall, R. (1992). Substance use and risky sexual behavior for exposure to HIV: Issues in methodology, interpretation, and prevention. Amercian Psychologist, 48, 1035-1045.

Leigh, B. C., Temple, M. T., & Trocki, K. F. (1994). The relationship of alcohol use to sexual activity in a U. S. national survey. Social Science and Medicine, 39, 1527-1535.

MacDonald, T. K., Fong, G. T., Zanna, M. P., & Martineau, A. M. (2000). Alcohol myopia and condom use: Can alcohol intoxication be associated with more prudent behavior? Journal of Personality and Social Psychology, 78, 605-619.

MacDonald, T. K., Zanna, M. P., & Fong, G. T. (1996). Why common sense goes out the window. Effects of alcohol on intentions to use condoms. Personality and Social Psychology Bulletin, 22, 763-775.

Marlatt, G. A., & Gordon, J. R. (Eds.). (1985). Relapse prevention: Maintenance strategies in the treatment of addictive behaviors. New York: Guilford Press.

Marlatt, G. A., & Rohsenow, D. J. (1980). Cognitive processes in alcohol use: Expectancy and the balanced-placebo design. In N. K. Mello (Ed.), Advances in substance abuse: Behavioral and biological research (Vol. 1, pp. 159-199). Greenwich, CT: JAI Press.

Martin, C. M., & Hoffman, M. A. (1993). Alcohol expectancies, living environment, peer influence and gender: A model of college-student drinking. Journal of College Student Development, 34, 206-212.

Maticka-Tyndale, E., & Herold, E. S. (1999). Condom use on spring-break vacation: The influence of intentions, prior use, and context. Journal of Applied Social Psychology 29, 1010-1027.

McEwan, R. T, McCallum, A., Bhopal, R. S., & Madhok, R. (1992). Sex and the risk of HIV infection: The role of alcohol. British Journal of Addiction, 87, 577-584.

Mckay, M. T. (1999). Mastery and alcohol expectancies as predictors of high-risk sexual behaviors in a population of females beginning alcohol treatment. Dissertation Abstracts International: Section B: The Sciences and Engineering, 59, 4450.

Meilman, P. W. (1993). Alcohol-induced sexual behavior on campus. College Health, 42, 27-38.

Morrison, D. M., Gilmore, M. R., & Baker, S. A. (1995). Determinants of condom use among high-risk heterosexual adults: A test of the theory of reasoned action. Journal of Applied Social Psychology, 25, 651-676.

Murphy, D. A., Rotheram-Borus, M. J., & Reid, H. M. (1998). Adolescent gender differences in HIV-related sexual risk acts, social-cognitive factors and behavioral skills. Journal of Adolescence, 21, 197-208.

Nicklin, J. L. (1999, May 28). Colleges report increases in arrests for drug and alcohol violations. The Chronicle of Higher Education, 44, A39.

O'Hare, T. (1999). Risky sex and drinking context in young women and men. Journal of Human Behavior in the Social Environment, 2, 1-18.

Oliver, M. B., & Hyde, J. S. (1993). Gender differences in sexuality: A meta-analysis. Psychological Bulletin 114, 29-51.

Raj, A. (1996). Identification of social cognitive variables as predictors of safer sex and intent in heterosexual college students. Journal of Sex and Marital Therapy, 22, 247-258.

Reisberg, L. (1998, September 18). More students abstain from alcohol and drink heavily, study finds. The Chronicle of Higher Education, 44, A49.

Roehrich, L., & Kinder, B. N. (1991). Alcohol expectancies and male sexuality: Review and implications for sex therapy. Journal of Sex and Marital Therapy, 17, 45-52.

Sheeran, P., & Orbell, S, (1998). Do intentions predict condom use? Meta-analysis and examination of moderator variables. British Journal of Social Psychology, 37, 231-250.

Sneed, C. D., & Morisky, D. E. (1998). Applying the Theory of Reasoned Action to condom use among sex workers. Social Behavior and Personality, 26, 317-328.

Stacy, A. (1997). Memory activation and expectancy as prospective predictors of alcohol and marijuana use. Journal of Abnormal Psychology, 106, 61-73.

Temple, M. T, & Leigh, B. C. (1992). Alcohol consumption and unsafe-sexual behavior in discrete events. The Journal of Sex Research, 29, 207-219.

Tolman, E. C. (! 932). Purposive behavior in animals and men. New York: Century.

Vasilias, J. G. (1999). Investigating the effects of message content, source and mode of presentation on AIDS-related attitudes and intentions. Dissertation Abstracts International: Section B: The Sciences and Engineering, 60, 0877.

Vogel-Sprott, M., & Fillmore, M. T. (1999). Expectancy and behavioral effects for socially used drugs. In I. Kirsch (Ed.), How expectancies shape experience (pp. 233-262). Washington, D.C.: American Psychological Association.

Wasserheit, J. (1997, September). STD prevention in the United States. National objectives and strategies. Paper presented at Workshop of Setting National STD Goals, Ottawa, Canada.

Weatherburn, P., Davies, P. M., Hickson, F. C., & Hunt, A. J. (1993). No connection between alcohol use and unsafer-sex among gay and bisexual men. AIDS, 7, 115-119.

Joseph W. LaBrie
Loyola Marymount University

Jason Schiffman and Mitchell Earleywine
University of Southern California

Adress correspondence to Joseph LaBrie, PO Box 45041, Los Angeles, CA 90045; email: [e-mail].