Female gambling behaviour: a case study of realist description

[摘要] Gambling is a complex social behaviour. How behaviour is shaped can vary within different historical and cultural contexts: to date, it is rare for the impact of these different contexts to be examined. The study of gambling has been (largely) entrenched within a bio-medical paradigm, where problematic gambling is viewed as an innate characteristic of the individual. This focus limits understanding about the ways in which gambling behaviour is shaped and also limits the range of policy responses to intervention with ‘problematic’ individuals. Specific examination of the way different contexts and mechanisms, both proximate and distal, shape behaviour has not been undertaken. The term ‘prisoners of the proximate’ (Hanlon et al, 2012) is an apt description of much contemporary gambling research.This thesis seeks to explore alternative ways to frame the study of gambling behaviour and argues that a focus on contexts and how behaviour varies for whom and under what circumstances is appropriate. This builds on Pawson and Tilley’s (1997) principles of realist evaluation and Pawson’s (2006) work on realist review to consider what realist description might look like as a form of empirical investigation. This includes recognition of the inherent subjectivity of all research and advocates an expansive analytical approach whereby many different types of evidence are brought together to examine a particular issue. To do this, this thesis draws on secondary analysis of existing data, historical evidence and theoretical review.This approach is applied to the study of female gambling behaviour. By drawing together data generated from the 1940s to the present day, it demonstrates how patterns of gambling behaviour are gendered and how gambling preferences vary based on prevailing social and political norms and legislation. This thesis argues that a process of ‘re-”feminisation”’ of gambling is evident in Britain today. In addition, the diversity of female gambling behaviour among different groups of women is explored, as is variation based on individual, social and spatial characteristics. This is achieved by using many different sources of data (mainly large-scale government surveys such as the Health Survey for England, the British Gambling Prevalence Survey series, the Taking Part survey) but also by supplementing these datasets with administrative information about the spatial patterning of gambling venues to broaden the scope of investigation. A number of different analytic techniques are used (factor analysis, latent class analysis, survival analysis and more standard descriptive methods) to explore how behaviour varies for different women in different circumstances. Using an expansive approach to secondary data analysis, whereby information from different studies is used to explore female patterns of behaviour from different viewpoints, creates a more nuanced understanding of female gambling behaviour. This is the purpose of realist description. It is an approach which recognises that not everything is the same for all people in all circumstances. Recognising this diversity at the outset of investigation provides a platform to explore this in depth. This thesis argues that this recognition should underpin the design and analysis of primary survey research to provide a more solid basis upon which to consider why behaviour varies. Doing so creates a solid foundation for a more considered examination of what type of policy interventions are most appropriate, for whom, and under what circumstances.

[发布机构] University:University of Glasgow;Department:School of Social and Political Sciences

[关键词] Gambling, realism, female