207 General Discussion Chapter 4, findings captured by the instrument concern only the subgroup influenced by the differences in the leniency of long-term care assessors. Importantly, obtaining LATEs is not necessarily a disadvantage: they might be particularly informative for policy purposes, as in these two studies. In Chapter 3, age 18 matters for triggering the transitional gap, when there is an abrupt change in several aspects of the youngster’s life. In Chapter 4, the applicants influenced by leniency are those at the margin of being (or not) considered eligible in the assessment process. These applicants would also likely be the most impacted by policies focusing on supported housing eligibility, such as changes in funding that could entail a shift in the eligibility threshold. More and more diverse evidence to strengthen our conclusions The introduction of this thesis outlined the need for quasi-experimental designs to generate policy-relevant evidence. The evaluation of quasi-experiments is based on quantitative methods and often leads to exclusively quantitative studies. This is a limitation to the policy relevance of the findings, which could be strengthened through mixed methods studies. Formally including a qualitative data component to complement the quantitative design would be beneficial in several ways, without any prejudice of the robustness required for the quantitative evaluation piece. First, as with any other quantitative study, qualitative data might be useful to generate and target the research hypotheses ahead of data collection and methodological decisions [27]. Second, quasi-experiment-specific, qualitative data should be systematically collected to inform the intervention allocation. Understanding this allocation, commonly related to the operational characteristics of the program or policy being evaluated, is at the core of the quasi-experiment robustness. Its investigation is often done based on ad-hoc anecdotal information available to the researcher informally and not reported in the papers. A qualitative component in a mixed methods study would strengthen the assessment of the assumptions behind quasi-experimental designs, which can often not be empirically tested. Additionally, collecting qualitative data after obtaining the quasi-experiment results would be of value to clarify the channels by which the policy or program led to the outcomes observed. Quasi-experiments themselves offer limited insight into these channels [27]. Such insight would be particularly relevant in the case of mental health inequalities by SES, in which psychosocial pathways might play a more significant role than material ones. SES impacts constructs such as identity and social position, influencing psychological resources such as self-efficacy, optimism or connectedness [28]. A very intuitive example is the one of unemployment. It leads to a loss of income and limitations to buy material necessities. Still, it also leads to a psychosocial process of “loss of self-esteem and feelings of worthlessness that affect health via direct psychobiological processes or through modified behaviours and lifestyles” [29]. Fully understanding how the psychosocial pathways contribute to health disparities might be challenging due to the nature of these processes. But it is also essential to guide effective strategies to 7
RkJQdWJsaXNoZXIy MjY0ODMw