Towards a method for synthesizing diverse evidence using hypotheses as common language

Abstract

Combining the findings obtained by different research methods in mixed-research synthesis could potentially contribute to a broader, more diverse evidence base for interventions. In this article we focus on the methodological challenges involved in synthesizing various types of research findings. We propose a method that uses hypotheses to facilitate the comparison and integration of such different findings. The method consists of four steps: (1) synthesizing findings per source of evidence, (2) formulating a mono-method hypothesis for each source, (3) integrating the monomethod hypotheses into one overall hypothesis, and (4) evaluating, using empirical data, whether the overall hypothesis better fits the data than each of the mono-method hypotheses. Using quantitative studies, qualitative studies and experts’ views in the substantive case of children and trauma, we will illustrate the proposed method. We conclude that the method provides a viable perspective for constructing an elaborate model that captures the knowledge from complementary sources.

Appendix

To investigate the views of experts in the field of children and trauma, a purposive sample six Dutch mental health care professionals were interviewed. Participants were emailed for their cooperation. Each interview (performed by the first author) was held at the expert’s office and lasted about an hour. The interviews were audio-recorded and transcribed and field notes were made by the interviewer. The interview questions were open-ended and involved personal acquaintance, introduction about the research (including discussing confidentiality), and uncovering the determinants of PTSD development and their relative importance.

The transcribed interviews and the field notes were analyzed using computer software for qualitative analysis (QSR NVivo 8). The data were open-coded. Codes were discussed among the first two authors and agreed upon. Next, axial and selective coding were performed, resulting in the following themes: type of trauma (single or multiple trauma, human against human violence or natural disasters, amount of loss of control), severity (level of gravity of trauma experienced), ordering chaos (ability to reorganize chaos of feelings, experiences and memories), feelings (e.g., guilt, shame, loneliness, fear, anger), safety (sense of safety experienced), parenting (parents’ handling of child in terms of protection, help coping, exemplifying healthy reaction and availability), avoidance (ability to avoid situations reminding of traumatic event), child characteristics (age, temperament, cognitive/social intelligence, development and self-image), support (help offered and understanding shown by friends, family and community), trust (amount of faith in surrounding people), care after trauma (consolation received during/shortly after trauma), culture (collectiveness of community and it’s conventions of trauma), interpersonal relationships (interactions with friends, teachers and family).