208 Chapter 7 close the gap. In this thesis, qualitative data would have been of great value to excel the current findings towards more policy actionable conclusions: in Chapter 2, to explore patients´ and practitioners´ views on the mechanisms behind the inequalities in treatment outcomes; in Chapter 3, to inquire low-income youngsters or their parents on the decision to forego care; in Chapter 4, to characterize the group of compliers through the enquiry of the long-term care assessors and to help explain the unexpected findings of decreased labour participation of the applicants; and in Chapter 5 , to clarify the mechanisms behind the heterogeneous effects by levels of psychotherapy supply. Unfortunately, using mixed methods research is far from a common practice in academic research. Several reasons include the lack of institutional incentives to conduct this type of research, lack of wellrecognised journals to publish it, limited interaction between quantitative and qualitative researchers, and limited knowledge about each other’s disciplines. Implementing mixed methods studies is even more challenging in the context of PhD trajectories, with welldefined time horizons and publication objectives. Last, we could strengthen this thesis’ findings in their robustness, generalizability and policy relevance by complementing them with different types of evidence. For example, other pieces of primary research would allow the practice of triangulation, which consists of obtaining multiple estimates for the same intervention or policy using different study designs, analytical methods and data sources [30]. Second, by complementing the primary research conducted with secondary (evidence synthesis) and tertiary (evidence products) research, as outlined in the evidence creation funnel. Unfortunately, producing such a comprehensive evidence basis requires the collaboration and coordination of several researchers and disciplines, which is often challenging for junior researchers to lead in their PhD trajectory. The potential of causality as a continuum One of the objectives of this thesis was to conduct quasi-experimental research to strengthen the knowledge basis of policy on mental health inequalities, concluding about the causal effects of interventions, programs and policies that might close or widen the mental health gap. As a first step to achieving this aim thousands of hours were used to seek relevant and robust quasi-experiments. When a policy with potential would be identified, the following steps included collecting basic information about the implementation and allocation of exposure, determining whether the existing data would allow studying the outcomes of interest and exploring the identifying assumptions. The final confirmation of finding a suitable quasi-experiment would often just come after arranging access to the data and analysing it enough to discard major violations of these assumptions. During this process, dozens of relevant policies, programs and interventions were left behind. Many exposures do not provide the conditions to identify an acceptable control group and be evaluated as a quasi-experiment. There are, however, many others that
RkJQdWJsaXNoZXIy MjY0ODMw