Methodological problems related to alcohol research among Turks and Moroccans living in the Netherlands: Findings from semi-structured interviews

Objectives. To identify factors related to alcohol use among Turks and Moroccans living in the Netherlands. Furthermore, to reveal methodological problems related to research among Turks and Moroccans in general and to alcohol research among these groups in particular. Design. Individual face-to-face interviews were carried out with Dutch researchers (n = 9), Turkish and Moroccan (health) practitioners working in the field with Turks (n = 4) or Moroccans (n = 2), and members of the target population with a Turkish (n = 3) or a Moroccan background (n = 2). Furthermore, focus-group interviews were held with Turkish women (n = 4), Turkish men (n = 3), Moroccan women (n = 4) and Moroccan men (n = 3) working as health professionals. Results. Alcohol use seems prevalent particularly among second-generation Turks and Moroccans and is related to: upbringing, influence of peer groups, integration and the degree in which Islamic rules are practised. Written questionnaires seem more appropriate for second-generation Turks and Moroccans, because they have fewer language problems and are more familiar with Western bureaucratic society. However, both generations may prefer face-to-face interviews since both groups fear that 'written' answers about the sensitive subject 'alcohol use' may somehow become known among community members. Similarly, an interviewer with a Dutch background may elicit more reliable answers about alcohol use than an interviewer with a Turkish or Moroccan background. Conclusion. In alcohol research special attention should be paid to second-generation Turks and Moroccans. Although it is probably easier to conduct alcohol studies in this group than in first-generation Turks and Moroccans, quantitative research is needed to test the hypothesis that written questionnaires elicit more reliable answers about alcohol use than face-to-face interviews. Furthermore, the influence of ethnic matching on response and data quality should be tested further.