Health Education Research, Vol. 19, No. 1, 15-28,
February 1, 2004
© 2004 Oxford University Press
What determines future smoking intentions of 12- to 13-year-old UK African-Caribbean, Indian, Pakistani and white young people?
1 School of Education and 2 Department of Public Health and Epidemiology, University of Birmingham, Birmingham B15 2TT, UK, 3 Department of Epidemiology and Health Sciences, University of Manchester, Manchester M13 9PT, UK, 4 Department of Preventive Medicine, Faculty of Medicine, University of Oviedo, 33006 Oviedo, Spain and 5 Department of Cancer Prevention and Health Promotion, Maastricht University, 6200 MD Maastricht, The Netherlands 6 Correspondence to: W. Markham, School of Health and Social Studies, University of Warwick, Coventry CV4 7AL, UK. e-mail: wolfgang.markham{at}warwick.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionReferences |
|---|
It is important to know when designing adolescent smoking interventions how ethnicity and gender influence intention. This paper reports an investigation into how ethnicity influences the smoking intentions of disadvantaged UK African-Caribbean (n = 275), Indian (n = 397), Pakistani (n = 687) and white (n = 1792) 12–13 year olds. The Attitudes–Social influences–Efficacy (ASE) model underpinned the study. It states that ASE determinants (advantages, disadvantages, social acceptance, social norms, modelling, perceived pressure and self-efficacy) directly influence behavioural intention. External factors (country, ethnicity and gender) indirectly influence intention by influencing ASE determinants. ASE determinant scores and future smoking intentions were measured. Linear regression analyses showed that smoking intention varied by ethnicity and gender. Differences in ASE scores largely explained these variations. Ethnicity and gender did not modify the predictive effects of equivalent ASE determinant scores on intention. Being a white boy had a small independent direct influence on intention, which was ascribed to affective beliefs underpinning fitness and sporting prowess. Otherwise, ethnicity had no independent direct effects on intention. Culturally appropriate interventions that aim to change cognitions underpinning ASE determinants and, thus, ASE scores would, consequently, be expected to be equally effective amongst disadvantaged UK African-Caribbean, Indian, Pakistani and white adolescents.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
Recent UK studies indicate that from the age of 13, smoking is more common amongst adolescent females than males (Goddard and Higgins, 2000
). However, in some black and ethnic minority communities the reverse is true, and smoking is more common amongst adolescent males than females. Nazroo et al. (Nazroo et al., 2000) reported that when everybody in their sample (11–15 year olds) was included, 3% of females and 2% of males were self-reported regular smokers. However, the proportion of females and males who smoked varied amongst the African-Caribbean, Indian and Pakistani participants [African-Caribbean females (3%) and males (1%); Indian females (0%) and males (3%); Pakistani females (0%) and males (0%)] (Nazroo et al., 2000). The observations of Nazroo et al. (Nazroo et al., 2000) were produced from approximately 100 participants per ethnic group, so they may be quite imprecise estimates of the true population prevalence. However, Kemm and Harvey (Kemm and Harvey, 1994) reported that although Asian adolescent females were less likely to smoke than Asian adolescent males, the reverse trend was observed amongst African-Caribbean and white adolescents. While Markham et al. (Markham et al., 2001) reported that smoking was much more common amongst 14- to 15-year-old Bangladeshi males (35%) than amongst 14- to 15-year-old Bangladeshi females (11%).
It may be deduced from the findings of Nazroo et al. (Nazroo et al., 2000) that smoking prevalence is lower amongst adolescents from some black and minority ethnic communities than amongst adolescents from the white communities. Most investigations into the reasons why adolescents smoke have focused on white adolescents. Generaliza tions from these studies have led to the development of interventions that are aimed at the typical white adolescent. It might be expected that smoking might play different roles and have different meanings for young people from black and minority ethnic communities, and that smoking prevention interventions aimed at the typical white adolescent may not, therefore, always be applicable. Thus, understanding why the prevalence of smoking varies according to both ethnicity and gender is not only inherently interesting, but may also be of practical importance.
A number of social cognition or psychosocial models purport to explain behaviour under voluntary control, such as smoking, including the Attitudes–Social influences–Efficacy (ASE) model (De Vries et al., 1988, 1995; De Vries and Backbier, 1994). The ASE model has its origins in the Theory of Reasoned Action (Fishbein and Ajzen, 1975) and Bandura’s Social Cognitive Theory (Bandura, 1976), and is closely related to the Theory of Planned Behavior (Ajzen, 1991). More recently, aspects of the Transtheoretical Model (Prochaska et al., 1992) have been incorporated into the ASE model (De Vries and Mudde, 1998). According to the ASE model, future health related behaviour is a function of current behavioural intention. Current behaviour, on the other hand, is determined by past behavioural intention. Behavioural intention regarding behaviour in the future is solely determined by three types of psychosocial mediating factors: attitudes, social influences and self-efficacy. Attitudes refer to expected outcomes of taking up the behaviour, subdivided into advantages and disadvantages of smoking. Social acceptance is a distinct subgroup of advantages and focuses on the perceived ability of smoking to facilitate social interactions. Social influences are comprised of social norms, modelling and perceived pressure. Social norms are participants’ perceived expectations of important others regarding the uptake of smoking by the participant. Modelling refers to perceived prevalence of smoking among influential people. Perceived pressure is experience of direct pressure to smoke. Thus, social influences may be direct (social norms, perceived pressure) or indirect (modelling). Self-efficacy is a person’s beliefs in her/his ability to behave in the way that she/he wishes to behave in respect of smoking. The components of the psychosocial mediating factors are known as the ASE determinants. In total, then, there are seven ASE determinants that directly influence behavioural intention regarding future behaviour: advantages, disadvantages, social acceptance, social norms, modelling, perceived pressure and self-efficacy. The ASE model postulates that other variables, such as country of residence, socioeconomic and demographic characteristics, are external to the theoretical model. These external factors only influence the seven types of ASE determinants, and, consequently, have an indirect rather than a direct influence on adolescent smoking intention and, thus, future behaviour. Evans et al. (Evans et al., 1988) maintain that these external factors only serve to characterize individuals and may, therefore, be classified as descriptive variables. The descriptive variables are, consequently, distinguishable from the ASE determinants.
The ASE model has been used in cross-sectional and longitudinal studies to explain adolescent smoking uptake (De Vries et al., 1995) and adult smoking cessation (Willemsen et al., 1996; De Vries and Mudde, 1998). The ASE model also underpinned two successful adolescent smoking prevention trials (De Vries et al., 1992, 1994; Dijkstra et al., 1999). The aim of interventions such as the ones in these trials is to change the cognitions underpinning the ASE determinants, which in turn alters behavioural intention and, thus, the likelihood of smoking in the future.
This paper reports on an investigation into how ethnicity influences the smoking intentions of disadvantaged UK adolescents (12–13 years old) from the African-Caribbean, Indian, Pakistani and white communities. The ASE model underpinned the investigation. As far as we know, no reported UK study has used a well-established underpinning explanatory framework to investigate the influences of ethnicity and gender on young peoples’ smoking intentions. This information could be used to develop culturally appropriate health education interventions.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
Sampling and data collection
The data were obtained from a cross-sectional study conducted in, 1997 in the Midlands, UK. All mainstream secondary schools serving disadvantaged communities in three urban areas in the Midlands were invited to take part in the study. Schools were eligible if at least 30% of pupils were entitled to free school meals, which indicates parents/guardians receive financial assistance from the State. Of the 45 eligible schools, 32 (71.1%) agreed to take part. Questionnaires were administered by teachers to all Year 8 pupils (12–13 years old) in each participating school who were present on the day the questionnaires were administered (n = 3716, 84.5% of registered pupils). Teachers were supplied with administration instructions and made aware of the importance of adhering to the protocol. Pupils were assured their answers were confidential and no teachers would see their completed questionnaires. They were also asked not to put their names on the questionnaires and to seal completed questionnaires in provided envelopes. There were 11 self-reported ethnicity categories (African-Caribbean, Bangladeshi, black African, black other, Chinese, Indian, Pakistani, Vietnamese, white, mixed parentage and other). Only people who reported they were African-Caribbean (n = 275), Indian (n = 397), Pakistani (n = 687) or white (n = 1792) were included in the data analysis (n = 3151, 84.8%). Pupils who were not African-Caribbean, Indian, Pakistani or white were excluded from the analysis, as there were insufficient numbers for meaningful analysis.
Questionnaire
The questionnaire focussed on participants’ demographic characteristics, socioeconomic circumstances, attitudes towards smoking, smoking social influences, self-efficacy with respect to smoking, smoking intention and smoking behaviour. Modood et al. (Modood et al., 1997) showed that the meaning of socioeconomic indicators varies according to ethnicity. For example, owner-occupied accommodation belonging to Pakistani families is worse, on average, than most rented accommodation (Modood et al., 1997). Thus, multiple markers of material disadvantage/advantage were required so that we could take into account confounding by socioeconomic status on the relationships between ethnicity and smoking. Four questions characterized socioeconomic status: free school meal entitlement, housing tenure, access to a car and sharing a bedroom. The questions on the ASE model were developed in The Netherlands by the originators of the ASE model (De Vries and Kok, 1986; De Vries et al., 1988, 1995; Dijkstra et al., 1999) and translated. The questionnaire was piloted informally with six groups of five mixed-ability Year 6 (10–11 years old) pupils to ensure pupils with relatively low academic attainment levels were able to interpret the questions appropriately. It was also piloted formally with a different mixed-ability Year 6 class (10–11 years old) to ensure it could be completed within the allotted time frame (40 min). The questions on intention and the ASE determinants were of the Likert-type scale with intervals of 1 (Table I). Positive scores translate as promoting non-smoking. For each participant, the scores of the individual questions pertaining to each distinct ASE determinant were added together and divided by the number of questions. Missing answers were counted as having the mean score for each separate question. Each participant, consequently, had an overall score for each ASE determinant.
|
Adolescents regularly smoking at least one cigarette a week were considered regular smokers (Bewley et al., 1972
; Leventhal and Cleary, 1980). Smoking status was assessed through the standard UK closed question developed by Bewley et al. (Bewley et al., 1972) and a dichotomous question that asked about participants’ recent smoking history. The standard closed question asked participants to describe themselves as never smokers, ex-smokers, occasional smokers or regular smokers who regularly smoke at least once a week or at least once a day. The dichotomous question asked ‘Have you smoked during the past 7 days?’. Pupils who were inconsistent in their answers to the two smoking status questions were excluded from the data analysis (n = 118: 3.7%).
Data analysis
There are three routes through which ethnicity could influence intention. First, as predicted by the ASE model, ethnicity could influence the distribution of ASE determinant scores. In other words, the influence of ethnicity is mediated through the ASE determinants. Second, contrary to the predictions of the ASE model, ethnicity could influence intention directly and apparently independently of the ASE determinants. This second route would suggest that cultural factors are not currently captured by the ASE determinants. Third, contrary to the predictions of the ASE model, ethnicity could modify the influence of equivalent ASE determinant scores on intention. All three routes were tested using linear regression analyses. All the analyses were performed using SPSS for Windows version 10.0. In all the linear regression models reported below, we examined for violation of the assumptions underlying multivariable regression. These were multivariable normality, i.e. normally distributed residuals, and homoscedasticity. No evidence of violations was observed.
Does intention to smoke vary by ethnicity?
We tabulated the means and SD of the ASE determinants and intention to smoke by ethnicity and gender. Variations in intention to smoke by ethnicity could arise from two sources. First, ethnicity could influence intention. Second, socioeconomic or demographic factors associated with intention to smoke could also be associated with ethnicity and hence confound apparent ethnic differences, either overstating or understating ethnicity’s influence on intention. To examine the influence of ethnicity on intention net of confounding, we conducted a linear regression analysis with intention as the outcome, included ethnicity as an independent variable, and adjusted for age, gender and markers of socioeconomic status (bedroom sharing, housing tenure, free school meal eligibility and household access to a car). All variables except age were entered as dummy terms. However, as discussed above, the meaning of socioeconomic status markers, such as housing tenure, varies by ethnicity (Modood et al., 1997). To allow the significance of each marker of socioeconomic status to vary by ethnic group, we created ethnicity x socioeconomic status marker interaction terms and adjusted for these in this and all subsequent analyses. These effects are not reported as they are not germane to the aims of this paper.
The literature reviewed in the Introduction implied that the effects of ethnicity might be different for boys and girls. We examined for this by creating ethnicity x gender interaction terms. These were entered as a block and the improvement in model fit was tested using the F-test.
Why is there an association between ethnicity and intention?
Having ruled out confounding by socioeconomic and demographic factors for the association of ethnicity with intention, we then examined two reasons why ethnicity might influence intention. These were:
• The effects of ethnicity on intention were mediated by the ASE determinants or, alternatively, a direct unmediated path existed between ethnicity and intention.
• The effects of ethnicity modified the influence of equivalent ASE determinant scores on intention.
The effects of ethnicity on intention were mediated by the ASE determinants
Three steps are involved in testing for mediation of the effect of ethnicity on intention through the ASE variables (Baron and Kenny, 1986; MacKinnon, 1994). Each of the following criteria should be satisfied to show mediation.
Step I Ethnicity influences intention
Step II Ethnicity influences each ASE determinant
Step III Each ASE determinant influences intention when controlling for ethnicity
Step I is the analysis described above under ‘Does intention to smoke vary by ethnicity?’. It is represented by Path 1 in Figure 1.
|
For Step II, we linearly regressed each ASE determinant on its own, in turn, on variables added in three steps. First, we controlled the potentially confounding variables of age, gender and socioeconomic status markers. In the second step, we added ethnicity and in the third step, we tested whether the effect of ethnicity depended upon gender by adding ethnicity x gender interaction terms. We used the F-test for each of the latter two steps to test the improvement in the fit of the model. This is represented by Path 2 in Figure 1.
For Step III, we linearly regressed intention on four blocks of variables. In the first step, we controlled for the potential confounders (age, gender and socioeconomic status markers). In the second, we added an ASE determinant on its own. In the third, we tested the improvement in fit by adding ethnicity. In the fourth, we tested the improvement in fit by adding the ethnicity x gender terms. If neither ethnicity nor ethnicity x gender improved the fit of the model, this would be evidence that the ASE determinant potentially fully mediated the influence of ethnicity on intention. This is the case providing that the ASE determinant in question is significantly related to intention, having controlled for ethnicity and ethnicity x gender. This regression equation is represented by Paths 1 and 3 in Figure 1. We repeated this procedure for each ASE determinant.
Each ASE determinant was positively correlated with every other one, so that evidence that the effect of ethnicity was mediated by a particular ASE determinant might actually reflect the fact that the ASE determinant that was examined was correlated with the true mediating ASE variable. Alternatively, persistent direct unmediated effects of ethnicity could be present in this analysis because each ASE determinant was examined on its own and the correlation of ASE determinants is not perfect. Under these conditions, the effect of ethnicity could be mediated by other unexamined ASE determinants. To examine for these possibilities, we created a model predicting intention with all ASE determinants and potential confounders (age, gender and socioeconomic status markers) included. To examine for unmediated effects of ethnicity we then tested whether adding ethnicity followed by ethnicity x gender improved the fit of the model.
The effects of ethnicity modified the influence of each ASE determinant on intention
This is represented by Path 4 in Figure 1. To examine for this possibility, we linearly regressed intention (the outcome variable) on the potential confounding variables (age, gender and socioeconomic status markers and socioeconomic status markers x ethnicity interactions), ethnicity, ethnicity x gender and all the ASE determinants. We then examined whether ethnicity x each ASE determinant improved the fit of the model using the F-test. We added each ethnicity x ASE determinant interaction term in turn.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
Characteristics of the sample
The socioeconomic and demographic characteristics of the sample are recorded in Table II. The proportion of boys and girls was similar in each ethnic group (Table II). Overall the African-Caribbean and white samples had similar socioeconomic profiles. However, slightly more white participants had access to a car and were not entitled to a free school meal. The Indian sample was least disadvantaged, according to three socioeconomic indicators (free school meal entitlement, housing tenure and car access). However, Indian pupils were more likely to share a bedroom than both African-Caribbean and white pupils, but less likely to share a bedroom than Pakistani pupils. The socioeconomic indicators for the Pakistani adolescents appeared contradictory. On the one hand, more Pakistani pupils than African-Caribbean, Indian and white pupils were entitled to free school meals or shared a bedroom indicating greater relative disadvantage. On the other hand, compared to the African-Caribbean and white pupils, a relatively large proportion of Pakistani participants lived in owner-occupied accommodation or had access to a car indicating greater relative affluence.
|
The proportion of pupils who smoked is recorded in Table II. As expected, there were differences in smoking prevalence between ethnic groups. Overall, regular smoking was least common amongst the African-Caribbean pupils and most common amongst the white pupils. Amongst the African-Caribbean and white pupils, regular smoking was more common amongst girls than boys, but amongst Indian and Pakistanis pupils, regular smoking was less common amongst girls than boys.
Does intention to smoke vary by ethnicity?
There were relatively small differences in intention between ethnic groups among boys, but relatively larger differences among girls (Table III). White girls then Pakistani boys were the least anti-smoking of all ethnic-gender groups, while Pakistani girls, and Indian boys and girls were the most anti-smoking of all groups. The influences of ethnicity and gender on future intentions to smoke did not, consequently, reflect the influences of ethnicity and gender on current smoking behaviour (Tables II and III). The influences of ethnicity and gender on future smoking intentions were potentially confounded by demographic and socioeconomic differences. However, having controlled for these influences, both ethnicity (P < 0.001) and ethnicity x gender interactions (P < 0.001) statistically significantly improved the fit of the model, confirming that ethnicity and gender were significantly associated with intention to smoke. The pattern apparent in the unadjusted means was also largely confirmed in the linear regression analysis (results not shown).
|
Why is there an association between ethnicity and intention?
The effects of ethnicity on intention were mediated by the ASE determinants or, alternatively, a direct unmediated path existed between ethnicity and intention
There was strong evidence that all but two ASE determinants mediated the difference in intention between ethnic-gender groups. The disadvantages of smoking was not a mediator because it was not related to ethnicity (Table IV). The perceived social acceptance gained by young people through smoking was not a mediator because it was not significantly related to intention (Table V). There was limited evidence to support a direct, unmediated, path between ethnic-gender groups and intention. When all the ASE determinants were added together in the model, the addition of ethnicity did not improve the fit of the model, F = 1.7, (3, 2726), P = 0.16. However, adding ethnicity x gender significantly improved the fit of the model, F = 7.6, (3, 2723), P < 0.001.
|
|
The strength of the mediated and direct paths from ethnicity to intention are shown in Figure 2. Only the statistically significant paths have coefficients displayed. These coefficients are the difference in ASE scores between white girls (the reference group) and the ethnic-gender group in question. No coefficient for African-Caribbean boys and girls differed from those of white girls. In addition, disadvantages and social acceptance are not included because they failed the criteria for being mediators. It is apparent from Figure 2 that the stronger anti-smoking intention of Indian boys and girls and Pakistani girls arises mostly through mediation by social influences, particularly the modelling construct. Figure 2 also shows that the apparent difference in intention between white boys and white girls is mostly unmediated. The size of the coefficient (0.3) for the direct path from ethnicity to intention for white boys is the same as the difference in mean intention scores between white boys and white girls (Table III). No other direct unmediated effects of ethnicity x gender interactions were observed.
|
The effects of ethnicity modified the influence of each ASE determinant on intention
The F-test indicated that adding each ethnicity x ASE determinant interaction term in turn did not improve the fit of the model. There was therefore no evidence to suggest that ethnicity-gender groups determined the strength of influence of any ASE determinant on intention.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
This study has shown that intention to smoke in the future varies according to gender and ethnicity. It has also shown that the ASE determinants of intention vary with gender and ethnicity. Most of the variation of future smoking intention between gender and ethnic groups is explained by variation in ASE determinant scores within these groups. This study, thus, largely confirms the predictions of the ASE model and ethnicity and gender may, consequently, be classified as descriptive characteristics (Evans et al., 1988
). Therefore, in the UK, a strategy to reduce adolescent smoking based on the framework of the ASE model would be appropriate for boys and girls in all the ethnic groups studied.
The role of bias in this study
The aim of this investigation was to understand relationships between the ASE determinants and intention, and how ethnicity influences these relationships. Non-participation is a major concern when an investigation aims to identify population means, but is less of an issue when the aim of the study is to understand the relationships between variables. Nevertheless, the participation rate in this study was high.
Using multiple questions to assess intention would have improved the reliability of our measure of intention. However, in common with most other studies that have investigated the predictive ability of the ASE model, closely related Theory of Reasoned Action and closely related Theory of Planned Behaviour, we used a single-item measure of intention (Sutton, 1998).
Our measure of intention is perhaps closer to behavioural expectation than behavioural intention, which could influence the validity of our measure of behavioural intention. However, Randall and Wolff (Randall and Wolff, 1994) and Sheeran and Orbell (Sheeran and Orbell, 1998) both concluded in their meta-analyses that behavioural intention is a fairly robust construct and that the type of measure of behavioural intention does not greatly influence the predictive ability of behavioural intention.
The influence of ethnicity and gender on the ASE determinants
The mediation effect of ethnicity and gender was greatest with respect to perceived social norms and modelling as judged by the size of the ethnicity coefficients in Table IV. In other words, the greatest influences of ethnicity were mediated through participants’ perceived expectations of important others regarding the uptake of smoking by the participant and participants perceived prevalence of smoking among influential people. These influences of ethnicity are likely to arise because shared cultural norms in some black and minority ethnic communities may make smoking less acceptable, and because smoking is less common in some black and minority ethnic communities. These findings would indicate that smoking interventions that aim to reduce the general acceptability of smoking in society and, thus, alter cultural norms rather than interventions that focus solely on adolescents are likely to have the greatest influence on smoking behaviour of the adolescents in our sample. There is empirical evidence to support this view. Perry et al. (Perry et al., 1992) found that the most effective adolescent smoking interventions had a community-wide focus.
The direct and independent influence of being a white boy on intention
Contrary to the predictions of the ASE model, the effect of ethnicity on intention for white boys was not influenced by adjustment for the ASE determinants suggesting that the ASE model does not currently capture some cultural factors specific to this group. This may reflect the residual effects of incomplete or inaccurate measurement of the ASE determinants and/or intention. However, these observations may be real. If they are real, what is their basis? A possible cause arises through the influence of affective beliefs on attitudes towards smoking, which were not measured in this study. Connor and Armitage (Connor and Armitage, 1998) distinguish between the influences of instrumental, affective and moral beliefs on attitudes. Even though the ASE model has subsequently developed to include rational and emotional attitudinal components (De Vries et al., 2000), our study focussed entirely on instrumental beliefs. As our study did not assess the influences of the participants’ affective salient beliefs concerning attitudes towards smoking, it is possible that we have not sufficiently measured the influences of attitudes on intention especially amongst the white young men. Lloyd and Lucas (Lloyd and Lucas, 1998) investigated largely white adolescents, and showed that issues to do with fitness and sporting prowess were protective factors against smoking amongst adolescent males. The salient beliefs concerning fitness and sporting prowess amongst white boys in our sample, may, therefore, focus on how these young people perceive that they would feel if they chose to smoke, and this affected fitness and sporting prowess and, thus, be grounded in feelings such as shame, guilt and regret.
Implications for future UK adolescent anti-smoking interventions
The ASE model is an explanatory model and the findings of this study could inform the development of future UK adolescent anti-smoking interventions. However, this study was cross-sectional so, although we have implied that the ASE determinants caused intention, we can only say that they were associated with intention in the way that would be predicted from the causal theory. If the findings from this cross-sectional study are borne out in prospective studies in the UK, as they have elsewhere (De Vries et al., 1998), then successful interventions could be developed from the ASE model (De Vries et al., 1992, 1994; Dijkstra et al., 1999). Ethnicity did not modify the influence of equivalent ASE determinant scores on intention nor did ethnicity have independent direct effects on intention. Consequently, interventions that aim to change cognition’s underpinning all the ASE determinants, except social acceptance would be expected to be equally effective amongst UK adolescents who are African-Caribbean, Indian, Pakistani or white. However, as highlighted by De Vries (De Vries, 1998), even if the objective is the same for every ethnic group, the nature of the interventions and their delivery need to be culturally specific in order to change the ASE scores. Having said this, studies in the US suggest that similar interventions with similar deliveries may be equally appropriate for adolescents from black and minority ethnic communities and from white communities. Botvin et al. (Botvin et al., 1989, 1992) and Ellickson and Bell (Ellickson and Bell, 1990) reported that American adolescent smoking prevention interventions based on psychosocial models that are similar to the ASE model were generalizable to disadvantaged African-American, Hispanic and white pupils.
|
Conclusion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
One of the most pressing public health issues facing economically developed countries is that adolescent smoking continues unabated and in some countries is even increasing (Jarvis, 1997
; Kraft and Svendsen, 1997; Hill et al., 1998; Redmond, 1999). Equally important is that relatively poor people are several times more likely to smoke than affluent people and are also less successful when they try to stop (Jarvis and Wardle, 1999). Our sample was recruited from schools serving disadvantaged communities and in some cases schools serving very disadvantaged communities. The findings from this study indicate that UK strategies to reduce adolescent smoking based on the ASE model would be appropriate for disadvantaged boys and girls in all the ethnic groups we studied. |
Acknowledgements |
|---|
We wish to acknowledge the contribution of Jenny Douglas who worked directly on this project from 1996 to 1999. We are grateful to the students and teachers who participated in the project. The European Commission—Directorate General V (98/CAN 32466) funded the project.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
Ajzen, I. (1991) The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211.[CrossRef][Web of Science]
Bandura, A. (1976) Social Foundations of Thought and Action: A Social Cognitive Theory. Prentice-Hall, Englewood Cliffs, NJ.
Baron, R.M. and Kenny, D.A. (1986) The moderator–mediator distinction in social psychological research. Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182.[CrossRef][Web of Science][Medline]
Bewley, B.R., Day, I. and Ide, L. (1972) Smoking by Children in Great Britain. A Review of the Literature. Social Science Research Council/ Medical Research Council, London.
Botvin, G.J., Batson, H., Witts-Vitale, S., Bess, V., Baker, E. and Dusenbury, L. (1989) A psychosocial approach to smoking prevention for urban Black youth. Public Health Reports, 104, 573–582.[Web of Science][Medline]
Botvin, G.J., Dusenbury, L., Baker, E., James-Ortiz, S. and Botvin, E.M. (1992) Smoking prevention among minority youth: assessing effects on outcome and mediating variables. Health Psychology, 11, 290–299.[CrossRef][Web of Science][Medline]
Connor, M. and Armitage, J.A. (1998) Extending the theory of planned behavior; a review and avenues for further research. Journal of Applied Social Psychology, 28, 1429–1464.[CrossRef]
De Vries, H. (1998) Planning and evaluating health promotion. In Scott, D. and Weston, R. (eds), Evaluating Health Promotion. Stanley Thomas, Cheltenham, pp. 92–108.
De Vries, H. and Backbier, E. (1994) Self efficacy as an important determinant of quitting among pregnant women who smoke; the omega pattern. Preventive Medicine, 23, 167–174.[CrossRef][Web of Science][Medline]
De Vries, H. and Kok, G. (1986) From determinants of smoking behaviour to the implications for a prevention program. Health Education Research, 1, 85–94.[Medline]
De Vries, H. and Mudde, A.N. (1998) Predicting stage transitions for smoking cessation applying the attitude–social influences–efficacy model. Psychological Health, 13, 369–385.
De Vries, H., Dijkstra, M. and Kuhlman, P. (1988) Self-efficacy the third factor besides attitude and subjective norm as a predictor of behavioural intention. Health Education Research, 3, 273–282.
De Vries, H., Dijkstra, M. and Kok, G. (1992) A Dutch smoking prevention program: and overview. Hygie, 11, 14–18.[Medline]
De Vries, H., Backbier, E. and Kok, G. (1994) A Dutch social influence smoking prevention approach for vocational school students. Health Education Research, 9, 365–375.
De Vries, H., Backbier, E., Kok, G. and Dijkstra, M. (1995) The impact of social influences in the context of attitude, self-efficacy, intention and previous behaviors as predictors of smoking onset. Journal of Applied Social Psychology, 25, 237–257.[Medline]
De Vries, H., Mudde, A.N., Dijkstra, M. and Willemsen, M.C. (1998) Differential beliefs, perceived social influences and self efficacy expectations among smokers in various motivational phases. Preventive Medicine, 27, 681–689.[CrossRef][Web of Science][Medline]
De Vries, H., Mudde, A. and Dijkstra, A. (2000). The Attitude–Social Influence–Efficacy-model: An application to the prediction of motivational transitions in the process of smoking cessation. In Norman, P., Abraham, C. and Conner, M. (eds), Understanding and Changing Health Behaviour: From Health Beliefs to Self-regulation. Harwood, New York, pp. 165–187.
Dijkstra, M., Mesters, I., De Vries, H., Van Breukelen, G. and Parcel, G.S. (1999) Effectiveness of two Dutch smoking prevention programmes with boosters. Health Education Research, 14, 791–802.
Ellickson, P.L. and Bell, R.M. (1990) Drug prevention in junior high. A multi-site longitudinal test. Science, 247, 1299–1305.
Evans, R.I., Dratt, L.M., Raines, B.E. and Rosenberg, S. (1988) Social influences on smoking initiation; importance of distinguishing descriptive versus mediating process variables. Journal of Applied Social Psychology, 18, 925–943.[CrossRef]
Fishbein, M. and Ajzen, I. (1975) Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research. Addison-Wesley, Reading MA.
Goddard, E. and Higgins, V. (2000) Drug Use, Smoking and Drinking among Young Teenagers in 1999. Office for National Statistics, London.
Hill, D.J., White, V.M. and Scollo, M.M. (1998) Smoking behaviours of Australian adults in 1995: trends and concerns. Medical Journal of Australia, 168, 209–213.[Web of Science][Medline]
Jarvis, L. (1997) Smoking among Secondary School Children in 1996: England. The Stationery Office, London.
Jarvis, M.J. and Wardle, J. (1999) Social patterning of individual health behaviours: the case of cigarette smoking. In Marmot, M. and Wilkinson, R.G. (eds), Social Determinants of Health. Oxford University Press, Oxford, pp. 240–255.
Kemm, J. and Harvey, J. (1994) Smoking in Birmingham School Children Junior Pulse: Report of Survey 1994. ‘Smoke Free Birmingham’ Project, Birmingham.
Kraft, P. and Svendsen, T. (1997) Tobacco use among young adults in Norway, 1973–95: has the decrease levelled out? Tobacco Control, 6, 27–32.[Abstract]
Leventhal, E.M. and Cleary, P.D. (1980) The smoking problem: a review of research and theory in behavioural risk modification. Psychological Bulletin, 88, 370–405.[CrossRef][Web of Science][Medline]
Lloyd, B. and Lucas, K. (1998) Smoking and Adolescence: Images and Identities. Routledge, London.
Markham, W.A., Featherstone, K., Taket, A., Trenchard-Mabere, E. and Ross, M. (2001) Smoking amongst UK Bangladeshi adolescents aged 14–15. Health Education Research, 16, 143–156.
MacKinnon, D.P. (1994) Analysis of mediating variables in prevention and intervention research. In Cazares, A. and Beatty, L.A. (eds). Scientific Methods in Prevention Research. Research Monograph 139, DHHS publ. no. 94–3631. NIDA, Washington, DC, pp. 127–153.
Modood, T., Berthoud, R., Lakey, K., Nazroo, J., Smith, P., Nirdee, S. and Beishon. S. (1997) Ethnic Minorities in Britain: Diversity and Disadvantage. Policy Studies Institute, London.
Nazroo, J., Becher, H., Kelly, Y. and McMunn, A. (2000) Children’s health. In Erens, B., Primatesta, P. and Prior, G. (eds), The Health Survey for England—The Health of Minority Ethnic Groups ’99. Vol. 1: Findings. The Stationary Office, London, pp. 345–395.
Perry, C., Kelder, S. and Knut-Inge, K. (1992) Community wide smoking prevention: long-term outcomes of the Minnesota Heart Health Programme and the Class of 1989 study. American Journal of Public Health, 82, 1210–1216.
Prochaska, J.O., DiClemente, C.C. and Norcross, J.C. (1992) In search of how people change: applications of addictive behaviours. American Psychologist, 47, 1102–1114.[CrossRef][Medline]
Randall, D.M. and Wolff, J.A. (1994) The time interval in the intention–behaviour relationship: meta analysis. British Journal of Social Psychology, 33, 405–418.
Redmond, W.H. (1999) Trends in adolescent cigarette use: the diffusion of daily smoking. Journal of Behavioral Medicine, 22, 379–395.[CrossRef][Web of Science][Medline]
Sheeran, P. and Orbell, S. (1998) Do intentions predict condom use? Meta analysis and examination of six moderator variables. British Journal of Social Psychology, 37, 231–250.[Web of Science][Medline]
Sutton, S. (1998) Predicting and explaining intentions and behavior: how well are we doing? Journal of Applied Social Psychology, 28, 1317–1338.[CrossRef][Web of Science]
Willemsen, M.C., De Vries, H., van Breukelen, G. and Oldenburg, B. (1996) Determinants of intention to quit smoking among Dutch employees: the influence of the social environment. Preventive Medicine, 25, 195–202.[CrossRef][Web of Science][Medline]
Received on January 31, 2002; accepted on November 20, 2002
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Panday, S. P. Reddy, R. A. C. Ruiter, E. Bergstrom, and H. de Vries Determinants of smoking among adolescents in the Southern Cape-Karoo region, South Africa Health Promot. Int., September 1, 2007; 22(3): 207 - 217. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Panday, S. P. Reddy, R. A. C. Ruiter, E. Bergstrom, and H. de Vries Determinants of smoking cessation among adolescents in South Africa Health Educ. Res., October 1, 2005; 20(5): 586 - 599. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||