Safe Sex

Abstract

Safe sex is a form of sexual behavior that cognitive psychology treats less as a hygiene instruction than as a chain of decisions made under risk, uncertainty, and competing motivation. The decision to use protection is first formed: attitudes, perceived social norms, and a sense of control combine into an intention, the central construct of the dominant behavior-change models. It is then tested in the moment, where sexual arousal and alcohol narrow attention to immediate reward and discount the delayed, probabilistic cost of infection, so an intention formed in a cool state can collapse in a hot one. And it is influenced from outside, by how risk information is framed and by interventions that aim to close the gap between what a person intends and what a person does. Three demonstrations trace intention formation, the heat-of-the-moment discount, and message framing.

Keywords: safe sex, decision-making, risk perception, theory of planned behavior, intention–behavior gap

Safe sex — the set of practices, condom use chief among them, that reduce the risk of sexually transmitted infection and unintended pregnancy — is usually taught as a list of recommendations. Read through cognitive psychology it becomes something more tractable: a sequence of decisions in which a person weighs a certain, immediate cost against an uncertain, delayed one, under the ordinary distortions that risk perception, social influence, and transient states impose. Each of those is a well-studied cognitive process, and each has been measured directly in the safe-sex literature rather than assumed (Sheeran et al., 1999; Albarracín et al., 2005).

Framing it this way neither moralizes nor lectures. It asks what is being decided, how the options are valued, what states and messages shift that valuation, and why an intention so often fails to become an act — the same questions cognitive science asks of any behavior. The influence runs both ways: a person's beliefs and intentions shape whether protection is used, and the experience of each encounter, in turn, updates the attitudes and perceived norms that the next decision starts from (Ajzen, 1991; Sheeran & Webb, 2016).

What Safe Sex Is

Safe sex, in the sense indexed by MeSH and used across the health-behavior literature, is the practice of sexual activity in ways that reduce the probability of transmitting a sexually transmitted infection or producing an unintended pregnancy — most centrally through barrier protection, but also through partner selection, testing, and negotiated agreements. MeSH files it as a form of sexual behavior, the parent descriptor under which it sits in the behavioral tree, reflecting that safe sex is not a separate activity but a manner of conducting an ordinary one. That placement matters: safe sex is a decision layered on a behavior, and the question cognitive psychology asks is how that decision is made, sustained, and sometimes abandoned (Sheeran et al., 1999).

The field's own history is a shift in what counts as the lever. Early campaigns assumed that supplying information would be enough — tell people the risk, and behavior will follow. Decades of intervention research replaced that assumption with a more structured account: information is necessary but weak on its own, and durable change requires motivation and concrete behavioral skills as well (Fisher & Fisher, 1992). The comprehensive meta-analytic record bears this out, with passive information campaigns producing far smaller effects than active interventions that build motivation and skill (Albarracín et al., 2005).

Three features make safe sex a revealing case for cognitive psychology. It is decided: whether to use protection is a choice valued under risk, shaped by beliefs about outcomes and by perceived norms (Ajzen, 1991). It is state-dependent: the same person decides differently when aroused or intoxicated, because those states change what the mind attends to (Ariely & Loewenstein, 2006; MacDonald et al., 1996). And it is fallible in execution: even a firm intention frequently fails to produce the act, a gap that is itself a target of theory and intervention (Sheeran & Webb, 2016).

How Safe Sex Intersects Cognition

Safe sex touches cognitive psychology at three points, each a different process operating on the same choice. First, intention formation: beliefs about outcomes, perceived social norms, and a sense of control combine into an intention to use protection. Second, state-dependent valuation: arousal and alcohol narrow attention and discount the delayed cost of infection, so the intention is tested under conditions that work against it. Third, influence and execution: how risk is framed, and how well an intention is translated into action, determine whether the intention becomes behavior. Table 1 lays out the three, and the three interactive demonstrations below take them in turn.

Table 1

Three Cognitive Problems Behind Safe Sex

Note. The three problems are facets of one choice, not stages that happen only once; a person forms an intention, has it tested by state, and is influenced by messages across every encounter. MeSH files Safe Sex (D021841) under Sexual Behavior (F01.145.802.845) and records no narrower descriptors beneath it, so the organizing structure here is cognitive rather than taxonomic.

Forming the intention: attitudes, norms, and control

The first cognitive problem is how a person arrives at an intention to use protection at all. The dominant answer is Ajzen's theory of planned behavior, which holds that behavior is driven proximally by an intention, and that intention is itself a weighted combination of three beliefs: the person's attitude toward the behavior (is using a condom good or bad, pleasant or unpleasant), the subjective norm (do important others approve), and perceived behavioral control (does the person feel able to do it, including negotiating it with a partner) (Ajzen, 1991). Applied to condom use, the model has strong empirical support: a classic meta-analysis found that attitudes, norms, and especially perceived control and communication predict condom use across dozens of studies, far better than knowledge of risk alone (Sheeran et al., 1999), and a more recent integrated meta-analysis in a high-incidence setting confirmed the same structure (Protogerou et al., 2018).

This is why the information deficit assumption — that people fail to use protection because they do not know the risk — has been so thoroughly displaced. Fisher and Fisher's information–motivation–behavioral-skills model makes the point directly: information is necessary but acts only through motivation and skill, so a person can be fully informed and still not protected if the motivation or the practical skill is missing (Fisher & Fisher, 1992). The broader meta-analytic record agrees, with the health belief model and protection motivation theory adding that perceived susceptibility, perceived severity, and the perceived efficacy of the protective response are what convert a threat into a motivation to act (Rosenstock, 1974; Rogers, 1975). The AIDS risk reduction model draws these threads into a sequence, holding that protective change proceeds in stages — first labeling a behavior as risky, then committing to change it, and finally enacting the change — each stage with its own cognitive determinants rather than a single global decision (Catania et al., 1990). Demo 1 makes the intention computation explicit: set the attitude, the perceived norm, and the perceived control, and the resulting intention — the weighted combination the model predicts — is drawn, showing how a weakness in any one determinant drags the intention down.

The heat of the moment: arousal and alcohol

The second problem is why an intention that is firm in a calm moment can evaporate in a charged one. The cleanest evidence is Ariely and Loewenstein's demonstration of the hot–cold empathy gap in sexual decision-making: the same men, asked about risky and protective choices once in a cool state and once in a state of sexual arousal, reported far greater willingness to engage in unprotected and risky behavior when aroused — and, crucially, underestimated that shift when cool, so they could not predict their own hot-state selves (Ariely & Loewenstein, 2006). Arousal does not merely raise motivation; it reweights the decision, inflating the immediate reward and shrinking the salience of a delayed, probabilistic harm.

Alcohol does something structurally similar through a different route. MacDonald, Zanna, and Fong showed that intoxication specifically lowers intentions to use condoms, and argued the mechanism is alcohol myopia: alcohol narrows attention to the most salient, immediate cues — here, arousal and a willing partner — and away from the abstract, distal cue of infection risk, so the risk simply drops out of the choice (MacDonald et al., 1996). This is not mere disinhibition; the systematic experimental evidence shows alcohol causally raises intended sexual risk-taking in a dose-dependent way consistent with the myopia account (Scott-Sheldon et al., 2016; Berry & Johnson, 2018). The common thread with arousal is temporal discounting: the farther off and less certain the harm, the less weight it carries, and both states steepen that discount. Johnson and Bruner's Sexual Discounting Task operationalizes exactly this, measuring how the likelihood of using a condom falls as the delay to, or the attractiveness of, a sexual reward rises (Johnson & Bruner, 2012). Demo 2 runs the trade-off: set a cool-state intention and then raise arousal and alcohol, and the effective intention is shown shrinking below the threshold at which protection is used — the mechanism behind the intention–behavior gap this state produces.

Changing behavior: framing and the intention–behavior gap

The third problem is how the decision can be shifted from outside, and why shifting it is harder than it looks. The influence lever best understood is message framing. Rothman and Salovey, applying prospect theory to health, showed that the persuasive power of a message depends on matching its frame to the behavior: gain-framed messages (emphasizing the benefit of acting) work better for prevention behaviors, which are perceived as safe and certain, while loss-framed messages (emphasizing the cost of not acting) work better for detection behaviors, which carry the risk of finding something wrong (Rothman & Salovey, 1997). Condom use is a prevention behavior, so the model predicts gain-framed appeals — protection preserves health and freedom — should outperform fear-based loss frames, a non-obvious consequence of how the same risk is cognitively packaged.

But framing a message well only moves the intention, and intention is only half the problem. The intention–behavior gap — the routine failure of a formed intention to produce the corresponding act — is large and well documented, with meta-analyses showing that a substantial change in intention yields only a modest change in behavior (Sheeran & Webb, 2016). This is why effective safe-sex interventions do not stop at persuasion: the meta-analytic evidence shows that changing attitudes, norms, and especially self-efficacy produces corresponding changes in health behavior, locating self-efficacy — the perceived control term of the belief model — as a particularly high-yield target (Sheeran et al., 2016). Framing and skill-building are complementary: one shapes the intention, the other helps the intention survive the journey to action. Demo 3 makes the framing effect visible: choose a behavior type and a message frame, and the predicted persuasion is shown rising when the frame matches the behavior and falling when it does not.

Worked Example

Model the condom decision to see why an intention formed in a cool moment can fail in a hot one. Represent the intention as the theory of planned behavior does — a weighted combination of three beliefs, each scaled from 0 to 1. Take a person with a favorable attitude A = 0.8, a moderate perceived norm N = 0.5, and fairly high perceived control C = 0.7, and weight the three equally. The intention is

I = (0.8 + 0.5 + 0.7) ÷ 3 = 0.67.

Suppose protection is used when the effective intention at the moment of choice is at least 0.5. In a cool state there is no discount, so the effective intention is 0.67, which clears the threshold — protection is used. Now move to the heat of the moment. Let arousal contribute a discount proportional to its level and alcohol contribute another, so the total discount is d = 0.5 × (arousal) + 0.5 × (alcohol), each on a 0-to-1 scale. With moderate-to-high arousal of 0.6 and mild intoxication of 0.2,

d = 0.5 × 0.6 + 0.5 × 0.2 = 0.40,

and the effective intention falls to

I_eff = 0.67 × (1 − 0.40) = 0.67 × 0.60 = 0.40.

The effective intention now sits below the 0.5 threshold, so the same person who intended to use protection does not — the intention–behavior gap, produced here by a state that discounts the delayed risk (Ariely & Loewenstein, 2006; MacDonald et al., 1996). Two readings follow. First, the failure is not a failure of knowledge or even of intention; the intention was adequate, and the state dismantled it. Second, the leverage points are visible in the arithmetic: raise the cool intention (a stronger attitude, norm, or sense of control pushes I well above 0.5 so it survives a discount), or reduce the discount (plan ahead, limit intoxication, make the protective act require no in-the-moment decision). Demo 2 runs exactly this computation, letting the cool intention, arousal, and alcohol vary so the threshold crossing can be watched directly, and Demo 1 shows how the cool intention itself is built from its three determinants.

Discussion

Reading safe sex through cognition does two useful things. It relocates the explanation of risky sex from character to process: if protection is a decision formed from beliefs, tested by transient states, and executed across a gap, then the failures are not evidence of recklessness but predictable outputs of ordinary cognitive machinery — the information deficit that is not really a deficit, the arousal that reweights the choice, the gap between intending and doing (Fisher & Fisher, 1992; Ariely & Loewenstein, 2006; Sheeran & Webb, 2016). And it sharpens intervention: the strongest levers are the ones the models name — build perceived control and self-efficacy, frame prevention as gain, and design around the hot state rather than assuming a cool one (Sheeran et al., 2016; Rothman & Salovey, 1997).

Two cautions keep the account honest. First, the belief models are powerful but incomplete: they predict intention well and behavior less well, which is precisely the gap that newer work addresses, and they say little about the relational and structural conditions — a partner's power, access to protection, economic pressure — that set the terms of any individual decision (Sheeran et al., 1999; Protogerou et al., 2018). Second, much of the sharpest experimental evidence on arousal and alcohol comes from specific samples and controlled settings, and translating a dose-dependent laboratory effect into a real encounter demands care (Scott-Sheldon et al., 2016; Berry & Johnson, 2018). The mechanisms are well supported; the full mapping from belief to state to act, across every population and context, is still being drawn.

Current Directions

The most active front is closing the intention–behavior gap rather than merely documenting it. Recent meta-analytic work pins down which mediators actually carry behavior change — finding that interventions moving self-efficacy and norms produce the largest downstream shifts — which turns the belief models from descriptions into design specifications for what an intervention must target (Sheeran et al., 2016; Sheeran & Webb, 2016). A second front develops the pharmacological account of risk, using controlled administration studies to separate alcohol's genuine causal effect on sexual decision-making from mere correlation, and to test the myopia mechanism against alternatives (Berry & Johnson, 2018; Scott-Sheldon et al., 2016). A third extends the integrated belief models into high-incidence and cross-cultural settings, where the relative weight of attitude, norm, and control shifts and the practical skill of negotiation becomes decisive (Protogerou et al., 2018). The question uniting them is whether protective behavior is best improved by changing what people believe, by designing for the state they will actually be in, or by building the specific skills that let an intention survive contact with the moment.

Common Misconceptions

People have unsafe sex because they do not know the risks.

The information-deficit assumption is largely wrong. Decades of intervention research show that information is necessary but weak on its own; a fully informed person will still go unprotected if motivation or the practical skill to act is missing. Durable change requires motivation and behavioral skills, not more facts (Fisher & Fisher, 1992; Albarracín et al., 2005).

Intending to use protection is basically the same as using it.

Intention and behavior are distinct, and the gap between them is large. Meta-analyses find that even a substantial change in intention produces only a modest change in behavior, which is why forming the intention is only half the problem and why interventions must also help the intention survive to the moment of action (Sheeran & Webb, 2016).

Alcohol leads to risky sex simply by removing inhibitions.

The better-supported account is alcohol myopia. Alcohol narrows attention to the most salient immediate cues and away from abstract, distal ones such as infection risk, so the risk drops out of the decision rather than being consciously overridden. The effect is specific and dose-dependent, not a blanket loss of restraint (MacDonald et al., 1996; Scott-Sheldon et al., 2016).

Scaring people with the worst outcomes is the best way to promote condoms.

Not for a prevention behavior. Message framing predicts that gain-framed appeals — emphasizing the benefit of protecting — outperform fear-based loss frames for prevention behaviors such as condom use, while loss frames suit detection behaviors such as screening. Matching the frame to the behavior matters more than the intensity of the fear (Rothman & Salovey, 1997).

Glossary

AIDS risk reduction model.

A stage model of protective change in which a person labels a behavior as risky, commits to changing it, and then enacts the change, each step with its own cognitive determinants (Catania et al., 1990).

Alcohol myopia.

The narrowing of attention under intoxication onto salient, immediate cues and away from distal, abstract ones, so that a delayed risk such as infection drops out of the decision (MacDonald et al., 1996).

Attitude.

A person's overall favorable or unfavorable evaluation of performing a behavior; one of the three determinants of intention in the theory of planned behavior (Ajzen, 1991).

Decision under risk.

A choice whose outcomes have known or estimable probabilities; the condom decision weighs a certain immediate cost against a probabilistic delayed one (Sheeran et al., 1999).

Gain-framed message.

A persuasive message that emphasizes the benefits of taking an action; predicted to be most effective for prevention behaviors such as condom use (Rothman & Salovey, 1997).

Health belief model.

An account in which protective action follows from perceived susceptibility, perceived severity, and the perceived benefits and barriers of acting (Rosenstock, 1974).

Hot–cold empathy gap.

The systematic failure to predict, while in a cool state, how one will decide in a hot (aroused or craving) state; it inflates the immediate reward and shrinks a delayed harm (Ariely & Loewenstein, 2006).

Information–motivation–behavioral-skills model.

A framework holding that information changes behavior only through motivation and concrete behavioral skills, so information alone is necessary but insufficient (Fisher & Fisher, 1992).

Intention–behavior gap.

The routine shortfall by which a formed intention fails to produce the corresponding act; a substantial change in intention yields only a modest change in behavior (Sheeran & Webb, 2016).

Intention.

A person's readiness to perform a behavior; in the theory of planned behavior it is the proximal cause of action and the weighted combination of attitude, norm, and perceived control (Ajzen, 1991).

Loss-framed message.

A persuasive message that emphasizes the costs of failing to act; predicted to be most effective for detection behaviors such as screening (Rothman & Salovey, 1997).

Perceived behavioral control.

A person's sense of being able to perform a behavior, including negotiating it with a partner; the determinant of intention most predictive of condom use and the highest-yield intervention target (Sheeran et al., 2016).

Protection motivation theory.

An account in which a threat produces motivation to protect only when the person also judges the protective response effective and themselves able to perform it (Rogers, 1975).

Safe sex.

The practice of sexual activity in ways that reduce the risk of infection or unintended pregnancy; cognitively, a decision formed from beliefs, tested by state, and executed across a gap (Sheeran et al., 1999).

Sexual behavior.

The parent category under which MeSH files safe sex; the broad class of behavior of which safe sex is a risk-reducing manner (Protogerou et al., 2018).

Sexual discounting.

The decline in the likelihood of using protection as the delay to, or the attractiveness of, a sexual reward rises; a form of temporal discounting measured by the Sexual Discounting Task (Johnson & Bruner, 2012).

Subjective norm.

A person's perception of whether important others approve of a behavior; one of the three determinants of intention in the theory of planned behavior (Ajzen, 1991).

Key Researchers

Icek Ajzen

(University of Massachusetts Amherst, Professor Emeritus of Psychology). Formulated the theory of planned behavior, the model that casts the intention to use protection as a weighted combination of attitude, subjective norm, and perceived behavioral control. ORCID · Google Scholar · Faculty page

Dolores Albarracín

(University of Pennsylvania, Annenberg School for Communication). Conducted the comprehensive meta-analysis of HIV-prevention interventions showing that active interventions building motivation and skill, not passive information, produce behavior change. Google Scholar · Faculty page · Wikipedia

Jeffrey D. Fisher

(University of Connecticut, Professor Emeritus of Psychological Sciences). Developed, with William Fisher, the information–motivation–behavioral-skills model, the field's dominant framework for why information alone does not change protective behavior. Faculty page · Wikipedia

Matthew W. Johnson

(Sheppard Pratt; formerly Johns Hopkins University School of Medicine). Created the Sexual Discounting Task, which operationalizes how the delay and attractiveness of a sexual reward erode the choice to use a condom. Google Scholar

George Loewenstein

(Carnegie Mellon University, Herbert A. Simon University Professor of Economics and Psychology). Demonstrated, with Dan Ariely, the hot–cold empathy gap in sexual decision-making, showing that arousal reweights the risk calculus in ways people cannot predict when cool. ORCID · Faculty page · Wikipedia

Paschal Sheeran

(University of North Carolina at Chapel Hill, Professor of Psychology and Neuroscience). Meta-analyzed the psychosocial correlates of condom use and the intention–behavior gap, establishing that forming an intention is necessary but far from sufficient for protective behavior. Google Scholar · Faculty page

Frequently Asked Questions

How is safe sex a decision-making problem rather than a matter of willpower?

Cognitive psychology treats the use of protection as a decision under risk: a choice weighing a certain, immediate cost — reduced spontaneity or sensation — against an uncertain, delayed one, the probability and severity of infection or unintended pregnancy. The theory of planned behavior models that decision as an intention built from three beliefs: attitude toward protection, perceived social norms, and perceived control over actually using it. Framing safe sex this way explains why willpower is the wrong lens: the decision can fail because the intention was never strong, because a transient state discounted the risk, or because the intention did not translate into action — three distinct failure modes, each with its own evidence-based remedy, none of which is simply trying harder.

Why do people who know the risks still have unprotected sex?

Because knowledge is only one weak input to the decision. The information-deficit assumption — that unsafe sex comes from not knowing the danger — has been displaced by evidence that information changes behavior only through motivation and concrete skills. A person can fully understand infection risk yet still go unprotected if the motivation is low, if negotiating protection with a partner feels impossible, or if a hot state has discounted the risk in the moment. This is why effective programs build motivation and behavioral skills rather than simply delivering facts, and why the best predictor of condom use is perceived control, not knowledge.

How do alcohol and arousal change the decision to use protection?

Both narrow attention toward immediate reward and away from the delayed, probabilistic cost of infection, so the risk carries less weight in the moment. Sexual arousal produces a hot–cold empathy gap: aroused decision-makers are far more willing to take risks than their own cool-state selves predict. Alcohol adds alcohol myopia, a narrowing of attention onto the most salient immediate cues that pushes abstract risk out of the decision entirely. The effect of alcohol is specific and dose-dependent rather than a blanket loss of restraint. In decision terms, both states steepen temporal discounting, which is why an intention formed calmly can collapse when the moment arrives.

What is the intention–behavior gap, and why does it matter for safe sex?

The intention–behavior gap is the routine shortfall by which a formed intention fails to produce the corresponding action. Meta-analyses show that even a large change in intention yields only a modest change in behavior, so intending to use protection is genuinely different from using it. It matters because many safe-sex campaigns aim only at intention — persuading people that protection is good — and then find behavior barely moves. Closing the gap requires additional tools: building self-efficacy, planning for the specific situation, and designing so the protective act requires as little in-the-moment decision as possible.

Does fear-based messaging actually make people practice safer sex?

Often less than a positive frame does. Research on message framing shows that gain-framed appeals, which emphasize the benefits of acting, outperform loss-framed, fear-based appeals for prevention behaviors such as condom use, whereas loss frames suit detection behaviors such as getting screened. Condom use is a prevention behavior, so a message stressing what protection preserves — health, freedom, peace of mind — is predicted to be more persuasive than one stressing the horror of infection. Matching the frame to the type of behavior matters more than the intensity of the fear, which can even backfire by prompting avoidance.

Which psychological theories best explain safe-sex behavior?

Several complementary models dominate. The theory of planned behavior explains how an intention forms from attitude, norms, and perceived control. The information–motivation–behavioral-skills model explains why information works only through motivation and skill. The health belief model and protection motivation theory add that a threat motivates action only when the person judges the protective response effective and themselves able to perform it, and the AIDS risk reduction model stages the whole process from labeling a behavior as risky through to enacting change. They converge on a shared message: perceived control and self-efficacy, not knowledge, are the decisive levers.

How does safe-sex behavior feed back into a person's future decisions?

The influence runs in both directions. Beliefs and intentions shape whether protection is used, and the outcome of each encounter — whether negotiating protection went well, whether a scare followed a lapse — updates the attitudes, perceived norms, and sense of control that the next decision starts from. A successful protected encounter raises self-efficacy and strengthens the intention next time; a coerced or awkward one can lower both. This is why protective behavior cannot be read from a single choice: it is a sequence in which each decision revises the beliefs that feed the next, and interventions that build early successes can set a self-reinforcing trajectory.

Can safe-sex decisions be improved by designing for the heat of the moment rather than against it?

Yes, and this is where current thinking is heading. Because arousal and alcohol reliably discount delayed risk, interventions that assume a cool, deliberate decision-maker tend to fail at the moment of choice. The more effective approach designs for the hot state: forming implementation intentions in advance that specify exactly when and how protection will be used, keeping protection immediately to hand so no deliberation is required, and building the negotiation as a rehearsed skill rather than an in-the-moment decision. Each tactic works by removing weight from the discounted deliberation and shifting it to a decision already made while cool.

References

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179–211. https://doi.org/10.1016/0749-5978(91)90020-T

Albarracín, D., Gillette, J. C., Earl, A. N., Glasman, L. R., Durantini, M. R., & Ho, M.-H. (2005). A test of major assumptions about behavior change: A comprehensive look at the effects of passive and active HIV-prevention interventions since the beginning of the epidemic. Psychological Bulletin, 131(6), 856–897. https://doi.org/10.1037/0033-2909.131.6.856

Ariely, D., & Loewenstein, G. (2006). The heat of the moment: The effect of sexual arousal on sexual decision making. Journal of Behavioral Decision Making, 19(2), 87–98. https://doi.org/10.1002/bdm.501

Berry, M. S., & Johnson, M. W. (2018). Does being drunk or high cause HIV sexual risk behavior? A systematic review of drug administration studies. Pharmacology Biochemistry and Behavior, 164, 125–138. https://doi.org/10.1016/j.pbb.2017.08.009

Catania, J. A., Kegeles, S. M., & Coates, T. J. (1990). Towards an understanding of risk behavior: An AIDS risk reduction model (ARRM). Health Education Quarterly, 17(1), 53–72. https://doi.org/10.1177/109019819001700107

Fisher, J. D., & Fisher, W. A. (1992). Changing AIDS-risk behavior. Psychological Bulletin, 111(3), 455–474. https://doi.org/10.1037/0033-2909.111.3.455

Johnson, M. W., & Bruner, N. R. (2012). The Sexual Discounting Task: HIV risk behavior and the discounting of delayed sexual rewards in cocaine dependence. Drug and Alcohol Dependence, 123(1–3), 15–21. https://doi.org/10.1016/j.drugalcdep.2011.09.032

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(8), 763–775. https://doi.org/10.1177/0146167296228001

Protogerou, C., Johnson, B. T., & Hagger, M. S. (2018). An integrated model of condom use in Sub-Saharan African youth: A meta-analysis. Health Psychology, 37(6), 586–602. https://doi.org/10.1037/hea0000604

Rogers, R. W. (1975). A protection motivation theory of fear appeals and attitude change. The Journal of Psychology, 91(1), 93–114. https://doi.org/10.1080/00223980.1975.9915803

Rosenstock, I. M. (1974). Historical origins of the Health Belief Model. Health Education Monographs, 2(4), 328–335. https://doi.org/10.1177/109019817400200403

Rothman, A. J., & Salovey, P. (1997). Shaping perceptions to motivate healthy behavior: The role of message framing. Psychological Bulletin, 121(1), 3–19. https://doi.org/10.1037/0033-2909.121.1.3

Scott-Sheldon, L. A. J., Carey, K. B., Cunningham, K., Johnson, B. T., & Carey, M. P. (2016). Alcohol use predicts sexual decision-making: A systematic review and meta-analysis of the experimental literature. AIDS and Behavior, 20(S1), 19–39. https://doi.org/10.1007/s10461-015-1108-9

Sheeran, P., Abraham, C., & Orbell, S. (1999). Psychosocial correlates of heterosexual condom use: A meta-analysis. Psychological Bulletin, 125(1), 90–132. https://doi.org/10.1037/0033-2909.125.1.90

Sheeran, P., Maki, A., Montanaro, E., Avishai-Yitshak, A., Bryan, A., Klein, W. M. P., Miles, E., & Rothman, A. J. (2016). The impact of changing attitudes, norms, and self-efficacy on health-related intentions and behavior: A meta-analysis. Health Psychology, 35(11), 1178–1188. https://doi.org/10.1037/hea0000387

Sheeran, P., & Webb, T. L. (2016). The intention–behavior gap. Social and Personality Psychology Compass, 10(9), 503–518. https://doi.org/10.1111/spc3.12265