Safer Sex and The Health Belief Model
Safer Sex and The Health Belief Model
Safer Sex and The Health Belief Model
To cite this article: Tanya L. Boone & Eva S. Lefkowitz (2004) Safer Sex and the Health Belief
Model, Journal of Psychology & Human Sexuality, 16:1, 51-68, DOI: 10.1300/J056v16n01_04
ABSTRACT. The goal of the present study was to build on the Health
Belief Model (HBM) by adding predictors of late adolescent safer sex
behavior: perceptions of peer norms for sexual behavior, and sexual atti-
tudes that emerge from socialization. Sexually active, late adolescent
college students (N = 154, 62.3% female; mean age 20.8 years, 76% Eu-
ropean American) participated in the study. Predictors from the original
HBM included perceived vulnerability, condom use self-efficacy, and
attitudes about condoms. In addition, peer norms for condom use and
sexual behavior, general sexual attitudes, and endorsement of the sexual
double standard were included as predictors of safer sex behavior. Atti-
tudes about condoms, perceived vulnerability, condom use self-effi-
cacy, and the sexual double standard emerged as significant correlates of
condom use. General sexual attitudes and the sexual double standard
were significantly correlated with alcohol use before or during sex. With
the addition of these variables, the regression models accounted for 28%
of the variance in condom use, and 14% of the variance in alcohol use be-
fore or during sex. [Article copies available for a fee from The Haworth Docu-
ment Delivery Service: 1-800-HAWORTH. E-mail address: <docdelivery@
Downloaded by [Athabasca University] at 04:19 20 June 2016
KEYWORDS. Health Belief Model, safer sex, condom use, alcohol use
The late adolescent, college years are some of the riskiest years in individu-
als’ lives. College students take health risks related to poor nutrition, sexually
transmitted infections, unplanned pregnancies, and drug/alcohol abuse (Na-
tional Institute on Alcohol Abuse and Alcoholism [NIAAA], 2002). The
risk-taking behaviors of this age group are related to developmental processes
of late adolescence and emerging adulthood, such as transformations in iden-
tity, changes in social relationships, and new, more independent living situa-
tions (Goldscheider & Davanzo, 1986; Whitbourne & Tesch, 1985). Two
health risks among this age group that have received a great deal of attention
are sexual risk behaviors, and alcohol use and abuse (NIAAA).
Individually and in combination, sexual behavior and alcohol use present
serious potential health risks. Presidents of U.S. colleges have identified binge
drinking as the number one problem on their campuses because of the negative
consequences associated with binge drinking (NIAAA, 2002). These conse-
quences include unplanned pregnancies, sexual violence, and infection with
sexually transmitted diseases. Previous research has demonstrated an associa-
tion between alcohol use and sexual risk taking (Graves, 1995). For example,
roughly half of the college students in one study indicated that they had en-
gaged in vaginal intercourse at least once and up to five times primarily be-
cause they had been drinking (Butcher, Manning, & O’Neal, 1991). A quarter
of those surveyed in another study indicated that they had consumed alcohol
before or during their most recent sexual encounter with a new partner
(MacNair-Semands & Simono, 1996). Further, adolescents who are intoxi-
cated when engaging in sexual activity are more likely to take risks such as not
using a condom (Jemmott & Jemmott, 1993; MacNair-Semands & Simono).
Given the associations between alcohol use and sexual risk taking, safer sex
was conceptualized for the present study as sexual activity in which a condom
was used and alcohol was not consumed before or during the sexual encounter.
Several theories and models have been applied to the alcohol-related and
sexual risk-taking behavior of late adolescents and young adults, including the
Health Belief Model (Brown, DiClemente, & Reynolds, 1991; Mahoney,
1995; Sands, Archer, & Puleo, 1998; Zimmerman & Olson, 1994). This model
predicts behavior and behavior change from individuals’ perceptions of sus-
Tanya L. Boone and Eva S. Lefkowitz 53
missing from the model, and that inclusion of these variables would enable the
model to explain more variance in adolescent safer sex behaviors (Brown et
al.; Lollis, Johnson, & Antoni, 1997). The goal of the present study was to
build on the Health Belief Model by including two additional predictors of ad-
olescent safer sex behavior: perceptions of peer norms related to sexual behav-
ior, and sexual attitudes that emerge from cultural and gender sexual
socialization.
can acquire and successfully use condoms, and that condoms will prevent
pregnancy and STIs, they are more likely to use condoms. Lack of self-effi-
cacy may act as a barrier to condom use if feelings of being unable to acquire
and use condoms outweigh the perceived benefits of condom use (Strecher et
al., 1997). In the present study, we assessed all of the components of the HBM
Downloaded by [Athabasca University] at 04:19 20 June 2016
cally, Lollis and his colleagues suggest that additional variance in sexual
behavior may be explained if additional benefits and barriers to safer sex that
are related to culture and socialization are included. For example, the general
sexual attitudes that individuals hold may act as barriers to safer sex. Previous
research indicates that late adolescents who hold liberal sexual attitudes tend
Downloaded by [Athabasca University] at 04:19 20 June 2016
METHOD
Participants
Downloaded by [Athabasca University] at 04:19 20 June 2016
Measures
Perceived vulnerability to AIDS. The 6-item Fear of AIDS subscale of the
Multidimensional AIDS Anxiety Questionnaire (Snell & Finney, 1996) was
used to assess the extent to which participants experience fear or concern in re-
gard to AIDS (e.g., “I feel scared when I think about catching AIDS from a
sexual partner”). Respondents rated their agreement with each item on a
5-point scale ranging from “strongly disagree” to “strongly agree.” This
measure has demonstrated good reliability in previous research (α = .85-.94;
Snell & Finney) and in the current study (α = .90). Our measure of perceived
vulnerability to AIDS assesses the susceptibility construct of the HBM.
Condom use self-efficacy. The Condom Use Self-Efficacy Scale was used
to assess participants’ self-efficacy for condom use (Basen-Engquist et al.,
Tanya L. Boone and Eva S. Lefkowitz 57
1996). The scale consists of three subscales (3 items each) that assess self-effi-
cacy for communication about condom use (e.g., “. . . how sure are you that
you could tell your partner that you want to start using condoms?”), self-effi-
cacy for buying and using condoms (e.g., “If you wanted to get a condom, how
sure are you that you could go to the store and buy one?”), and barriers to con-
Downloaded by [Athabasca University] at 04:19 20 June 2016
dom use (e.g., “I would feel uncomfortable carrying condoms with me”). For
the communication and buying/using subscales, respondents rated their ability
to enact the behavior on a 3-point scale, from “not at all sure” to “totally sure.”
For the barriers subscale, respondents rated their agreement with each state-
ment on a 5-point scale, from “strongly disagree” to “strongly agree.” Higher
scores on the communication and buying/using subscales indicate high
self-efficacy, and higher scores on the barriers subscale reflect lower self-effi-
cacy. For this study, all three subscales were combined to create a total condom
use self-efficacy measure. Basen-Engquist et al. (1996) reported adequate reli-
ability (α = .66-.73). In the current sample, reliability for the overall measure
was also adequate (α = .65). Our measure of condom use self-efficacy assesses
the self-efficacy construct of the HBM.
Outcome expectancies for condom use. The Outcome Expectancies for
Condom Use Scale (Jemmott & Jemmott, 1992) was used to measure partici-
pant attitudes about condoms. The first subscale of the measure consists of
three items and assesses preventive expectancies, or the extent to which partic-
ipants believe that condoms can protect them from pregnancy and STDs in-
cluding AIDS (e.g., “Condoms can prevent sexually transmitted diseases”).
The second subscale of the measure consists of five items that assess hedonis-
tic expectancies, or the extent to which participants hold positive or negative
attitudes about condoms (e.g., “Sex feels good when a condom is used”). Par-
ticipants rated their agreement with the statements in each subscale on a
5-point scale ranging from “strongly disagree” to “strongly agree.” Jemmott
and Jemmott reported adequate reliability for this scale (α = .50-.73). In the
current sample, reliability was good (α = .83 for prevention, α = .76 for hedo-
nistic). These measures represent the benefits construct of the HBM, and can
also be viewed as measures of the absence of barriers to condom use.
Peer norms for sexual behavior. To assess perceptions of their peers’ norms
for sexual behavior, participants were asked to respond to an 8-item measure
that assesses peer approval of specific sexual behaviors at four levels of rela-
tionship involvement (Treboux & Busch-Rossnagel, 1995). For example, par-
ticipants indicated the extent to which their peers would approve of behaviors
such as “Sexual intercourse with someone you have gone out with once or
twice.” Participants rated their peers’ approval on a 4-point scale ranging from
“would strongly disapprove” to “would strongly approve.” As with previous
research (α = .90; Treboux & Busch-Rossnagel), this measure demonstrated
good reliability with the current sample (α = .80).
Peer norms for condom use. The Peer Norms for Condom Use (Gomez &
Marin, 1996) scale asks participants to rate the condom use behaviors of their
58 JOURNAL OF PSYCHOLOGY & HUMAN SEXUALITY
friends (e.g., “How many of your peers think using a condom is a good
idea?”). An expanded, 6-item version was developed from the original mea-
sure. Participants indicated their perceptions of their peers’ norms for condom
use on a 5-point scale ranging from “almost none” to “almost all.” This mea-
sure demonstrated adequate reliability in previous studies (α = .64-.73;
Downloaded by [Athabasca University] at 04:19 20 June 2016
RESULTS
TABLE 1. Percent of Participants Having Sex Without Condoms and with Alcohol
Condom Use
To examine associations between the predictor variables and participants’
lifetime condom use, partial correlations were conducted, controlling for age
(see Table 3). Among the males in the sample, condom use related to per-
ceived vulnerability and to hedonistic outcome expectancies for condom use.
Specifically, more frequent condom use was associated with stronger fears of
AIDS and with more positive hedonistic expectations for condom use. Among
the females in the sample, more frequent condom use related to higher condom
use self-efficacy, positive hedonistic outcome expectancies for condom use,
and endorsement of the sexual double standard.
60 JOURNAL OF PSYCHOLOGY & HUMAN SEXUALITY
Peer Peer
Condom Outcome Outcome Norms– Norms– Sexual
Downloaded by [Athabasca University] at 04:19 20 June 2016
Sexual -- .31***
Attitudes
Sexual Double --
Standard
Note. Due to missing data, sample size ranged from N = 152 to 154.
* p < .05; ** p < .01; *** p < .001.
and socialization variables were entered in the third step. Frequency of lifetime
sexual intercourse with a condom was the dependent variable (see Table 4). In step
one, age was negatively associated with frequency of sex with a condom such
that younger participants reported more frequent condom use. In step two, in
addition to age and gender, hedonistic outcome expectancies for condom use
were related to lifetime frequency of condom use during sexual intercourse.
Those participants who reported positive attitudes about condoms reported
more frequent condom use. Finally, in step three, age and hedonistic outcome
expectancies were related to condom use just as in step two. Peer acceptance
of sexual behavior was also marginally associated with more frequent condom
use. In addition, the sexual double standard was associated with frequency of
condom use such that endorsement of the sexual double standard related to
TABLE 4. Regression Model Predicting Condom Use from HBM Variables Plus
Peer Norms and Sexual Socialization Variables (N = 150)
Variable B SE B  R2⌬
DISCUSSION
The present study supports recent contentions that the ability of the Health
Belief Model to predict late adolescent and young adult safer sex behaviors
TABLE 5. Partial Correlations Between Alcohol Use During Sexual Intercourse
and Predictor Variables (Controlling for Age)
R⌬
2
Variable B SE B ß
2
Dependent Variable: Lifetime alcohol use during sexual intercourse (R = .14)
Step 1 .02
Age .09 .06 .12
Gender ⫺.08 .21 ⫺.03
Step 2 .03
Age .08 .06 .12
Gender .08 .22 .03
Perceived Vulnerability ⫺.01 .02 ⫺.07
Condom Use Self-Efficacy .02 .03 .06
Outcome Expectancies–Prevention .05 .04 .12
Outcome Expectancies–Hedonistic ⫺.04 .03 ⫺.12
Step 3 .10**
Age .10 .06 .15
Gender ⫺.08 .23 ⫺.03
Perceived Vulnerability .00 .02 .00
Condom Use Self-Efficacy .01 .03 .03
Outcome Expectancies–Prevention .02 .04 .05
Outcome Expectancies–Hedonistic ⫺.05 .03 ⫺.15
Peer Norms–Sexual Behavior ⫺.05 .03 ⫺.01
Peer Norms–Condom Use .02 .03 .05
Sexual Attitudes ⫺.03 .01 ⫺.28**
Sexual Double Standard ⫺.02 .02 ⫺.12
sex is still pleasurable with condoms are more likely to use condoms. Impor-
tant for prevention and intervention planners is the consistent connection be-
tween the belief that sex is pleasurable when condoms are used and the use of
condoms. This finding suggests that a continued focus on condoms as protec-
tive as well as natural, and not an impediment to sexual pleasure, will encour-
age condom use among this age group.
It is important to note that positive attitudes about condoms were associated
with condom use but not with alcohol use before or during sex. This is not sur-
prising given that this variable specifically addresses sexual behavior. This
finding does, however, suggest important questions related to attitudes and ex-
pectations for sexual behavior when drinking. For this age group, it may be
important to understand attitudes about condom use specifically in situations
in which individuals have been drinking. For example, do individuals have
positive attitudes about using condoms in situations in which they have been
drinking? LaBrie, Schiffman, and Earleywine (2002) addressed a similar
question and found that among college men, alcohol use was associated with
the expectation that such alcohol use would impair decision-making abilities
and condom use skills. In turn, those expectancies related to a decreased inten-
tion to use condoms when drinking. Future efforts at predicting actual condom
use from condom use attitudes and expectancies when alcohol is involved will
be important for understanding the associations between condom attitudes and
condom use.
Perceived vulnerability to AIDS, one of the central predictors of the HBM
(Strecher et al., 1997), was only related to condom use for males. Males who
reported higher fear of AIDS reported more frequent lifetime condom use.
Fear of AIDS was not, however, related to lifetime condom use for females.
Although perceived vulnerability to AIDS was related to males’ condom use
in the partial correlations, it did not emerge as a significant predictor of con-
dom use in the regression. This finding suggests that among college students,
perceived vulnerability may not be as critical in predicting condom use as the
HBM posits. As suggested by others (Boone et al., 2003), the role of perceived
vulnerability in the HBM may need to be reconsidered in models predicting
sexual behaviors.
Condom use self-efficacy, another central variable in the HBM, also
emerged as a correlate of condom use, but only for females. As predicted by
the HBM (Strecher et al., 1997), self-efficacy in acquiring, communicating
about, and using condoms was positively associated in the present study with
lifetime condom use. Although self-efficacy was related to condom use in the
partial correlations, it was not a significant predictor of condom use in the re-
Tanya L. Boone and Eva S. Lefkowitz 65
gression. This finding suggests that at least for college students, condom use
self-efficacy may not contribute much to the prediction of actual condom use.
It is possible that as a result of receiving education and information about con-
doms, the individuals in this age group have such strong feelings of condom
use self-efficacy that self-efficacy does not distinguish the condom users from
Downloaded by [Athabasca University] at 04:19 20 June 2016
Women who endorse the sexual double standard, and therefore expect this be-
havior from their sexual partners, may be highly cognizant of the possibility of
being infected by their partners. To avoid these risks, these individuals who
endorse the sexual double standard may be more likely to use condoms. It is
clear that more research will be required to fully understand this association.
Downloaded by [Athabasca University] at 04:19 20 June 2016
Global sexual attitudes were related to alcohol use before or during sex, but
not to condom use. It is somewhat puzzling that sexual attitudes related to al-
cohol use but not to condom use. One possible explanation for this pattern is
the degree to which alcohol use and condom use are viewed as risky among
this age group. It appears that most college students view using condoms as ef-
fective risk prevention (Holtzman, Bland, Lansky, & Mack, 2001). In con-
trast, there is less evidence that the majority of college students recognize the
risks associated with using alcohol. Therefore, among those individuals who
are sexually active, their level of sexual conservatism may not be associated
with the choice to use condoms, in that most students perceive condoms as
protective. In contrast, more sexually conservative youth may refrain from al-
cohol use because they fear the loss of control, whereas more sexually liberal
youth may not have the same concerns, and therefore may use alcohol more
freely. Future research aimed at understanding the nuances of the associations
between sexual attitudes, alcohol use, and condom use will be helpful for un-
derstanding sexual risk in this age group, and for informing future interven-
tions.
It is important to note a number of limitations in this study. First, the use of
a college sample prevents us from generalizing these findings to individuals of
the same age who are not in college. For example, it is possible that non-col-
lege individuals are less influenced by peer norms. Second, our sample was
predominantly European American. Accordingly, we must be careful in
generalizing our findings to members of other ethnic groups. Finally, the
cross-sectional nature of our design prevents us from drawing any conclusions
about causality. We cannot conclude, for example, that outcome expectancies
for condom use, sexual attitudes, the sexual double standard, and peer norms
caused the individuals in our study to use or not use condoms or alcohol. As
Brunswick and Banaszak-Holl (1996) suggest, it is just as possible that the risk
behaviors are causing the attitudes and beliefs included in our model.
Despite these limitations, several important findings emerged from this
study. Although the Health Belief Model has been somewhat successful in pre-
dicting safer sex behaviors among late adolescents and emerging adults, it can
be improved upon. We have demonstrated that including peer norms and sexual
socialization variables improves the predictive ability of the model. Peer norms
for sexual behavior and condom use, sexual attitudes, and the sexual double
standard help us to understand late adolescent and emerging adult sexual behav-
ior within the context of the variables outlined in the Health Belief Model. In the
future, it will be important to examine these additional variables further to more
fully understand their contribution to the Health Belief Model.
Tanya L. Boone and Eva S. Lefkowitz 67
REFERENCES
Andrews, J. A., Tildesley, E., Hops, H., & Li, F. (2002). The influence of peers on
young adult substance use. Health Psychology, 21, 349-357.
Basen-Engquist, K., Masse, L., Coyle, K., Parcel, G. S., Banspach, S., Kirby, D., &
Downloaded by [Athabasca University] at 04:19 20 June 2016
Lollis, C. M., Johnson, E. H., & Antoni, M. H. (1997). The efficacy of the Health Belief
Model for predicting condom usage and risky sexual practices in university stu-
dents. AIDS Education and Prevention, 9, 551-563.
MacNair-Semands, R. R., & Simono, R. B. (1996). College student risk behaviors: Im-
plications for the HIV-AIDS pandemic. Journal of College Student Development,
Downloaded by [Athabasca University] at 04:19 20 June 2016
37, 574-583.
Madden, T. J., Ellen, P. S., & Ajzen, I. (1992). A comparison of the Theory of Planned
Behavior and the Theory of Reasoned Action. Personality and Social Psychology
Bulletin, 18, 3-9.
Mahoney, C. A. (1995). The role of cues, self-efficacy, level of worry, and high-risk
behaviors in college student condom use. Journal of Sex Education and Therapy,
21, 103-116.
Maiman, L. A., & Becker, M. H. (1974). The Health Belief Model: Origins and corre-
lates in psychological theory. In M. H. Becker (Ed.), The Health Belief Model and
personal health behavior. Thorofare, NJ: Charles B. Slack, Inc.
Muehlenhard, C. L., & Quackenbush, D. M. (1996). The social meaning of women’s
condom use: The sexual double standard and women’s beliefs about the meaning
ascribed to condom use. Unpublished manuscript.
National Institute on Alcohol Abuse and Alcoholism (2002). College drinking: Chang-
ing the culture.
Osgood, D. W., & Lee, H. (1993). Leisure activities, age, and adult roles across the life
span. Society and Leisure, 16, 181-208.
Sands, T., Archer, J., Jr., & Puleo, S. (1998). Prevention of health-risk behaviors in col-
lege students: Evaluation seven variables. Journal of College Student Development,
38, 331-342.
Smith, B. N., & Stasson, M. F. (2000). A comparison of health behavior constructs: So-
cial psychological predictors of AIDS-preventive behavioral intentions. Journal of
Applied Social Psychology, 30, 443-462.
Snell, W. E., Jr., & Finney, P. (1996). The Multidimensional AIDS Anxiety Question-
naire. Unpublished manuscript.
Steers, W. N., Elliott, E., Nemiro, J., Ditman, D., & Oskamp, S. (1996). Health beliefs
as predictors of HIV-preventive behavior and ethnic differences in prediction. The
Journal of Social Psychology, 136, 99-110.
Strecher, V. J., Champion, V. L., & Rosenstock, I. M. (1997). The Health Belief Model
and health behavior. In D. S. Gochman (Ed.), Handbook of health behavior re-
search I: Personal and social determinants (pp. 71-91). New York: Plenum Press.
Strecher, V. J., DeVellis, B. M., Becker, M. H., & Rosenstock, I. M. (1986). The role of
self-efficacy in achieving health behavior change. Health Education Quarterly, 13,
73-92.
Treboux, D., & Busch-Rossnagel, N. A. (1995). Age differences in parent and peer influ-
ences on female sexual behavior. Journal of Research on Adolescence, 5, 469-487.
Whitbourne, S. K., & Tesch, S. A. (1985). A comparison of identity and intimacy sta-
tuses in college students and alumni. Developmental Psychology, 21, 1039-1044.
Zimmerman, R. S., & Olson, K. (1994). AIDS-related risk behavior and behavior
change in a sexually active, heterosexual sample: A test of three models of preven-
tion. AIDS Education and Prevention, 6, 189-204.