As we have seen in earlier chapters, social inclusion is a complex concept which overlaps with several other important ideas in social and community psychiatry and which has potentially wide applicability to a range of individuals and groups.

In this chapter we will discuss some of the key methodological issues and research outcomes relating to research in social exclusion and mental ill health. Throughout, we will be concerned with research that addresses both social exclusion and mental illness, not with either alone, although they are clearly related. Given the breadth of this field, our review will be selective, aiming to give a general overview of the literature, rather than a comprehensive account.

Measurement

Conceptual and methodological issues

1. 1 individual-level measures reflect the situation and life experience of people belonging to socially excluded groups (in this case people with mental health problems); they are usually of more interest to mental health practitioners and clinically-oriented researchers

2. 2 societal-level measures, also sometimes referred to as a ‘social indicators’ approach, rely on existing data sources that characterise levels of participation in larger community samples – examples of this are employment rates as assessed in the annual Labour Force Survey (Department for Work and Pensions, 2007) and the number of rough sleepers as calculated by local authorities (Palmer et al, 2007); this approach tends to be of more interest to planners, politicians and social policy-makers.

Notwithstanding the importance of reliability, any measurement tool for social inclusion must also have validity. Thus, it must ‘converge’ with other instruments measuring similar concepts, such as poverty, citizenship, social capital, ‘quality of life’. However, it must also ‘diverge’ (discriminate) between social inclusion and these related concepts. For example, social inclusion combines aspects of poverty (e.g. income, welfare benefits, exclusion from the labour market) but also elements of social participation. Similarly, it includes aspects of social capital (e.g. membership of social networks, community groups and civic participation) but adds in poverty and stigma and discrimination. It also overlaps with quality of life measures in terms of both objective and subjective indicators. Any comprehensive measure of social inclusion must therefore meet these challenges to content and discriminant validity.