predictive of smoking behaviour (Murray et al. 1984; McNeil et al. 1988; Charlton
and Blair 1989; Gillies and Galt 1990; Goddard 1990). The main factor that predicts
smoking is parental smoking, with reports that children are twice as likely to smoke
if their parents smoke (Lader and Matheson 1991). In addition, parents’ attitudes to
smoking also influence their offsprings’ behaviour. For example, if a child perceives the
parents as being strongly against smoking, he or she is up to seven times less likely to be
a smoker (Murray et al. 1984). The next most important influence on smoking is peer
group pressure. Studies in the USA have examined the relationship between peer group
identity and tobacco use. The results showed that individuals who are identified by
themselves and others as being problem-prone, doing poorly at school, rarely involved in
school sports, high in risk-taking behaviour such as alcohol and drug use, and with low
self-esteem were more likely to have smoked (Mosbach and Leventhal 1988; Sussman
et al. 1990). On the other hand, research has also found that high rates of smoking
can also be found in children who are seen as leaders of academic and social activities,
have high self-esteem and are regarded as popular by their peers (Mosbach and Leventhal
1988). Another factor that influences whether children smoke is the attitude of their
school to smoking behaviour. A Cancer Research Campaign study (1991) found that
smoking prevalence was lower in schools that had a ‘no smoking’ policy, particularly
if this policy included staff as well as students. In summary, social factors such as the
behaviour and beliefs of parents, peers and schools influence the beliefs and behaviours
of children. Using an entirely different methodology, Graham used interviews with low
income women with pre-school children to explore the contextual factors which may
maintain smoking behaviour. She argued that although smoking is seen by researchers
as unhealthy and something to be prevented, the women in her study regarded smoking
as central to their attempts to ‘reconcile health keeping and housekeeping when their
reserves of emotional and physical energy may be seriously depleted’ (Graham 1987:
55). She stated that smoking works to promote these women’s sense of well-being and to
help them cope with caring. She reports that smoking can be seen as ‘the only activity
they do, just for themselves’. Smoking is therefore a product not only of beliefs but also
an individual’s social world.Alcohol initiation and maintenance
Most people try alcohol at some time in their lives. However, one survey showed that
about a quarter of those questioned (men and women) described themselves as lifelong
abstainers (GHS 1994). The most common reasons for never drinking alcohol were
religion and not liking it. Therefore, rather than examining predictors of drinking ‘ever’
or ‘occasionally’, this section examines what factors predict developing a problem with
drinking.Psychological predictors of alcohol initiation and maintenance
The tension-reduction hypothesis (Cappell and Greeley 1987) suggests that individuals
may develop a drink problem because alcohol reduces tension and anxiety. Tension
creates a heightened state of arousal and alcohol reduces this state, which perpetuates112 HEALTH PSYCHOLOGY