Social cognition models

Page 197
Black blue
SEX
197
studies suggest that unprotected anal sex is resurfacing as an increasing problem. This
suggests that the substantial increase in condom use which occurred after the initial
AIDS prevention efforts may be beginning to reverse.
Condom use is recommended as the main means to prevent the spread of the HIV
virus. These data suggest that many individuals do report using condoms, although
not always on a regular basis. In addition, many individuals say that they do not use
condoms. Therefore, although the health promotion messages may be reaching many
individuals, many others are not complying with their recommendations.
Predicting condom use
Simple models using knowledge only have been used to examine condom use. However,
these models ignore the individual’s beliefs and assume that simply increasing knowledge
about HIV will promote safe sex. In order to incorporate an individual’s cognitive state,
social cognition models have been applied to condom use in the context of HIV and AIDS
(see Chapter 2 for a discussion of these models). These models are similar to those used
to predict other health-related behaviours, including contraceptive use for pregnancy
avoidance, and illustrate varying attempts to understand cognitions in the context of the
relationship and the broader social context.
Social cognition models
The health belief model (HBM)
The HBM was developed by Rosenstock and Becher (e.g. Rosenstock 1966; Becker and
Rosenstock 1987) (see Chapter 2) and has been used to predict condom use. McCusker
et al. (1989) adapted the HBM to predict condom use in homosexual men over a 12month period. They reported that the components of the model were not good predictors
and only perceived susceptibility was related to condom use. In addition, they reported
that the best predictor was previous risk behaviour. This suggests that condom use is a
habitual behaviour and that placing current condom use into the context of time and
habits may be the way to assess this behaviour.
The reasons why the HBM fails to predict condom use have been examined by
Abraham and Sheeran (1994). They suggest the following explanations:
I Consensus of severity: everyone knows that HIV is a very serious disease. This
presents the problem of a ceiling effect with only small differences in ratings of this
variable.
I Failure to acknowledge personal susceptibility: although people appear to know about
HIV, its causes and how it is transmitted, feelings of immunity and low susceptibility
(‘it won’t happen to me’) are extremely common. This presents the problem of a floor
effect with little individual variability. Therefore, these two central components of the
HBM are unlikely to distinguish between condom users and non-users.
I Safer sex requires long-term maintenance of behaviour: the HBM may be a good predictor
of short-term changes in behaviour (e.g. taking up an exercise class, stopping
Page 197
Black blue
Page 198
Black blue
198 HEALTH PSYCHOLOGY
smoking in the short term), but safer sex is an ongoing behaviour, which requires an
ongoing determination to adopt condom use as a habit.
I Sex is emotional and involves a level of high arousal: these factors may make the rational
information processing approach of the HBM redundant.
I Sex is interactive and involves negotiation: condom use takes place between two people,
it involves a process of negotiation and occurs within the context of a relationship.
Assessing individual cognitions does not access this negotiation process. Abraham
and Sheeran (1993) suggest that social skills may be better predictors of safe sex.
I From beliefs to behaviour: the HBM does not clarify how beliefs (e.g. ‘I feel at risk’) are
translated into behaviour (e.g. ‘I am using a condom’).
The theory of reasoned action (TRA) and the theory of planned
behaviour (TPB)
In an attempt to resolve some of the problems with the HBM, the TRA and TPB have
been used to predict condom use. These models address the problem of how beliefs are
turned into action using the ‘behavioural intentions’ component. In addition, they
attempt to address the problem of placing beliefs within a context by an emphasis on
social cognitions (the normative beliefs component). Research suggests that the TRA has
had some degree of success at predicting condom use with behavioural intentions predicting condom use at one, three and four months (Fisher 1984; Boyd and Wandersman
1991; van der Velde et al. 1992). In addition, attitudes to condoms predict behavioural
intentions (Boldero et al. 1992), and perceived partner support (partner norms) appears
to be a good predictor of condom use by women (Weisman et al. 1991).
Research has also explored the relative usefulness of the TRA compared with the TPB
at predicting intentions to use condoms (Sutton et al. 1999). The results from this study
indicated that the TPB was not more effective than the TRA (in contrast to the authors’
predictions) and that past behaviour was the most powerful predictor. In a recent study
of condom use, the best predictors appeared to be a combination of normative beliefs
involving peers, friends, siblings, previous partners, parents and the general public.
This suggests that although cognitions may play a role in predicting condom use, this
essentially interactive behaviour is probably best understood within the context of both
the relationship and the broader social world, highlighting the important role of social
cognitions in the form of normative beliefs. Therefore, although TRA and TPB, address
some of the problems with the HBM, they still do not address some of the others.
The role of self-efficacy
The concept of self-efficacy (Bandura 1977) has been incorporated into many models of
behaviour. In terms of condom use, self-efficacy can refer to factors such as confidence in
buying condoms, confidence in using condoms or confidence in suggesting that condoms
are used. Research has highlighted an association between perceived self-efficacy and
reported condom use (Richard and van der Pligt 1991), and a denial of HIV risk during
the contemplation of sex (Abraham et al. 1994). Schwarzer (1992) developed the health
Page 198
Black blue
Page 199
Black blue
SEX
199
action process approach (HAPA, see Chapter 2), which places self-efficacy in a central
role for predicting behaviour. In addition, this model may be particularly relevant to
condom use as it emphasizes time and habit.
Problems with social cognition models
Many of the problems highlighted by the HBM are also characteristic of other social
cognition models (see Chapter 2 for a detailed discussion). These problems can be
summarized as follows:
1 Inconsistent findings. The research examining condom use has not produced consistent results. For example, whereas Fisher (1984) reported an association between
intentions and actual behaviour, Abraham et al. (1991) did not. Joseph et al. (1987)
suggested that condom use is predicted by peer norms, whereas Catania et al. (1989)
found that it is not. Furthermore, Catania et al. (1989) reported that condom use
relates to perceived severity and self-efficacy, but Hingson et al. (1990) said that it
does not relate to these factors. However, such studies have used very different populations (homosexual, heterosexual, adolescents, adults). Perhaps models of condom use
should be constructed to fit the cognitive sets of different populations; attempts to
develop one model for everyone may ignore the multitude of different cognitions held
by different individuals within different groups.
2 Sex as a result of individual cognitions. Models that emphasize cognitions and
information processing intrinsically regard behaviour as the result of information
processing – an individualistic approach to behaviour. In particular, early models
tended to focus on representations of an individual’s risks without taking into
account their interactions with the outside world. Furthermore, models such as the
HBM emphasized this process as rational. However, recent social cognition models
have attempted to remedy this situation by emphasizing cognitions about the individual’s social world (the normative beliefs) and by including elements of emotion
(the behaviour becomes less rational).
3 Perception of susceptibility. In addition, these models predict that because people
appear to know that HIV is an extremely serious disease, and they know how it is
transmitted, they will feel vulnerable (e.g. ‘HIV is transmitted by unprotected sex, I
have unprotected sex, therefore I am at risk from HIV’). This does not appear to be the
case. Furthermore, the models predict that high levels of susceptibility will relate to
less risk-taking behaviour (e.g. ‘I am at risk, therefore I will use condoms’). Again this
association is problematic.
4 Sex as an interaction between individuals – the relationship context. Models of
condom use focus on cognitions. In attempts to include an analysis of the place of
this behaviour (the relationship), variables such as peer norms, partner norms and
partner support have been added. However, these variables are still accessed by
asking one individual about their beliefs about the relationship. Perhaps this is still
only accessing a cognition not the interaction.
5 Sex in a social context. Sex also takes place within a broader social context,
involving norms about sexual practices, gender roles and stereotypes, the role of
Page 199
Black blue