adapting models to fit the specific factors related to a specific behaviour. This
study is interesting as it represents an attempt to integrate different models of
health behaviour.
Background
Since the identification of the HIV virus, research has developed means to predict
and therefore promote safer sexual behaviour. The PMT suggests that behaviour is a
consequence of an appraisal of the threat and an appraisal of the individual’s coping
resources. It suggests that these factors elicit a state called ‘protection motivation’,
which maintains any activity to cope with the threat. This study examines the role
of the PMT in predicting sexual behaviour and in addition examines the effect of
expanding the PMT to include variables such as coping styles, social norms and previous
behaviour.
Methodology
Subjects A total of 147 homosexual and 84 heterosexual subjects with multiple
partners in the past six months took part in the study. They were recruited from Amster-
dam through a variety of sources including informants, advertisements and a housing
service.
Design Subjects completed a questionnaire (either postal or delivered).
Questionnaire The questionnaire consisted of items on the following areas rated on a
5-point Likert scale:
1 Sexual behaviour and behavioural intentions: the subjects were asked about their
sexual behaviour during the previous six months, including the number and type of
partners, frequencies of various sexual techniques, condom use and future intentions.
2 Protection motivation variables: (a) perceived severity, (b) perceived susceptibility,
(c) response efficacy, (d) self-efficacy, (e) fear.
3 Additional beliefs: (a) social norms, (b) costs, (c) benefits, (d) knowledge, (e)
situational constraints.
4 In addition, the authors included variables from Janis and Mann’s (1977) conflict
theory: (a) vigilance, (b) hypervigilance, (c) defensive avoidance.
Results
The results were analysed to examine the best predictors of sexual behaviour in both
homosexual and heterosexual subjects. It was found that although the variables of the
PMT were predictive of behaviour and behavioural intentions in both populations, the
results were improved with the additional variables. For example, when social norms and
previous behaviour were also considered, there was improved associations with future
behaviour. In addition, the results suggested that although there was a relationship
HEALTH BELIEFS 29