Differences between perceived vulnerability and perceived risk Implications

Differences between perceived vulnerability and perceived risk: Implications for health theory and interventions Jennifer J. Harman, Ph. D Colorado State University

2005 -2010 Assistant Professor, Applied Social Psychology Colorado State University Remained an affiliate of CHIP Got married and had 2 children

Harman, J. J. , Wilson, K. , & Keneski, E. (2010). Social and environmental contributors to perceived vulnerability and perception of risk for negative health consequences. In J. G. Lavino & R. B. Neumann (Eds. ), Psychology of Risk Perception, pp. 1 -45. Hauppauge, NY: Nova Science Publishers, Inc.

Background Risk perception for HIV infection in intimate relationships • Harman, Smith & Egan, (2007) • Harman, O’Grady & Wilson (2009) Seemingly no differences in high risk versus lower risk populations • Harman, Wilson & Keneski (2010)

Background (cont. ) Information Behavioral Skills Behavior Motivation Adapted from Fisher & Fisher, 1992

Background (cont. ) Information Motivation Behavioral Skills Behavior Motivation Attitudes Social Norms Perceived Vulnerability

Perceived vulnerability (PV) versus Perception of risk (Po. R) Terms have been used interchangeably in health promotion/risk prevention literature Affect/feeling • “ I feel vulnerable to getting HIV” Cognitive/beliefs • “I think I am at high risk for getting HIV”

Now we know our ABCs… Affective attitudes Behavioral attitudes Cognitive attitudes

Two separate constructs Perceived Vulnerability (PV) Affective in nature Perception of Risk (Po. R) Cognitive in nature

Health Behavior Theories and PV Health Belief model (Rosenstock, 1974) Protection Motivation Theory (Rogers, 1983) Extended Parallel Process Model (Witte, 1992)

Why should I care? Research support for PV as a predictor of attitudes, intentions and outcomes is inconsistent. Simple health concerns: PV usually related • E. g. , adherence to a medical regimen following a sports injury Complex health concerns: less consistent • E. g. , genetic risk information for cancer

Development PV Classical conditioning & Po. R Linkages between acquired other automatic associative processes information and attitude object E. g. , fear-smoking E. g. , beliefs about exercise- diabetes Probability important

PV and Po. R and health outcomes Negative Relationship? Positive Relationship? Defensive behavior activation Protective behavior activation E. g. , PV + condom use Optimistic biases (e. g. , Lek & Bishop, 1995) Denial

So what is the problem? Health behavior change interventions often introduce threats to increase PV or Po. R If a defensive response is activated, this “threat” may backfire

The measurement bugaboo PV and Po. R measurements often combined or not reported PV: affective measures/automatic associations IAT, facial expression instruments, physiological reactions, cartoon face identification Po. R: cognitive measures of beliefs Self-report

The intervention challenge Interventions manipulate specific variables to create change in psychological and/or health outcomes Social and environmental contributors to PV and Po. R proximal in nature Social Environmental

Changing PV Implicit attitude change (Gawronski & Bodenhausen, 2001) • Change how associations are made • E. g. , associate a new feeling with the behavior • Social marketing • Change activation of pre-existing patterns of associations

Changing Po. R Explicit attitude change strategies Change in associative evaluation • Gradual change of associative patterns lead to change in Po. R Change in propositions relevant for judgments • E. g. , provide risk information Change in strategy to achieve consistency • E. g. , “It can happen to you” campaigns

Narrative Intervention Review Med. Line and Psychinfo lit search 936 Total Citations 90 “eligible” articles 59 studies remained after through review

Strategies used 76 intervention elements Vast majority targeted Po. R • 73% used second route of Po. R change • 15. 4% used third strategy (e. g. , cognitive dissonance) Only 8 interventions targeted PV • Used 1 st strategy Majority measured Po. R, consistent with what was targeted

A recent empirical example HIV disproportionately affects Blacks and Hispanics in the U. S. (CDC, 2008) Incarcerated populations 5 -6 times more likely to be infected than general population (Lopez et al. , 2001) Social antecedents of PV/Po. R? PV: past HIV risk behavior, past HIV testing Po. R: believe HIV is a problem in community, know someone who is infected

Research Qs Are PV and Po. R empirically distinct from one another? Would heterosexual individuals impacted by incarceration have higher levels of PV and Po. R than non-impacted individuals? Is PV higher with reports of past HIV risk behavior and less frequent HIV testing? Is Po. R higher when people believe HIV is a serious problem in their community and/or whether they know someone infected? Are there different relationships between the social antecedents of PV and Po. R for each sample? What is the relationship between PV and Po. R and attitudes towards condoms, intentions, and condom use?

Method Participants Two heterosexual couple samples • Impacted sample • Non-impacted sample Instruments PV: I don’t worry about HIV Po. R: It is really unlikely that I will get HIV PV determinants: • How often are you high on non-injected drugs or alcohol when you have sex? • How many times have you been tested for HIV? Po. R determinants: • How many people do you know who have or had HIV/AIDS? • How serious is HIV in your community? Condom Attitudes, Intentions and Use

RQs 1 & 2 RQ 1: Are PV and Po. R distinct? Correlations ranged from. 40 -. 67 for all samples RQ 2: Do impacted individuals have higher PV and Po. R? No! Males: reported less PV • t(101)= -2. 65, p =. 009 Males and females less Po. R • t (101) = -6. 77 men • t (101) = -5. 78 women • ps <. 001

RQ 3 & 5 Does being high in drugs or alcohol during sex influence PV? Did not influence PV, or Po. R Does previous HIV testing influence PV? Impacted sample tested much more frequently than nonimpacted sample Did not influence PV, or Po. R

RQ 4 & 5 Does the belief that HIV is serious problem in the community influence Po. R? Impacted sample saw it as a significantly more serious problem (ps <. 001) Not related to Po. R for any sample Belief lowered PV for non-impacted males! Does knowing someone who has/had HIV influence Po. R? Impacted sample knew more people Not related to Po. R for any sample Knowing someone lowered PV for non-impacted males

RQ 6 Condom Attitudes PV predicted more positive attitudes among impacted women and more negative attitudes among non-impacted women Intentions to use condoms PV predicted lower intentions to use among nonimpacted women Condom use Po. R for non-impacted women and impacted men associated with lower reports of condom use

Discussion of empirical example PV and Po. R are moderately related, but distinct PV and Po. R lower among impacted men and women Past risk behaviors and testing were not related to PV or Po. R Other antecedents operating?

Conclusion PV = affect/automatic associations Po. R= cognitive/explicit beliefs/propositions Different strategies and social/environmental determinants should be used to change them Measurement should reflect affective and cognitive aspects

Conclusion PV and Po. R should operate similarly across different negative health outcomes HIV, cancer, diabetes Considerable differences may exist between individuals and groups of differing risks Once differences are identified, explore reasons behind the differences, then develop tailored interventions E. g. , experimental testing of social/environmental determinants for change among specific groups

Future directions Create a valid measure of PV and Po. R In progress now Retest interventions that have manipulated PV and/or Po. R using new measure to determine if change occurs Manipulate external/situational cues to determine effect on PV and Po. R

Thanks! National Institute of Health #F 31 -MH 069079, a Grant-in-Aid from the Society for the Psychological Study of Social Issues, and a research grant from division 38 of the American Psychological Association (Health Psychology) Kristina Wilson & Liz Keneski Peter Mc. Graw, Hannah Gould, and Heather Patrick