Table 4.

Triangulation qualitative and quantitative data

Qualitative data	Quantitative data	Triangulation
Reasoning
Expectations based on theory.	Hypothesis 1 (H1) : Various forms of social capital are present in Ghana. (Identification of horizontal and vertical SC factors)
Iterative process engaging clients and integrating their emic perspectives on social relations, reciprocal support and trust.	Hypothesis 2 (H2) : Clients’ active membership in the NHIS is explained by social capital.
Results
(Individual interviews, FGD)	(Household survey)
Perceptions reveal existence of different forms of social capital.	4 factors identified: 2 horizontal, 2 vertical (see Table 2)	Consistent results: Both qualitative and quantitative methods identified different forms of social capital.
Horizontal: Traditionally strong family bonds. Multiple groups (youth, church, women), increasingly popular. Trust in traditional (informal) health system.	Horizontal:	The emic perspectives (qualitative) make findings more culturally relevant.
Vertical: Relationships of clients with healthcare providers and health insurer exist but are influenced by various issues.	- Trust (in community)	The identified forms of social capital in the interviews help quantify social capital (factors) in the survey
	- Solidarity & collective action (in community)
	Vertical:
	- Trust in HCP
	- Trust in NHIS
Clients value mutual support structures and groups for information sharing, motivation, communal action.	Multinomial logic regression of social capital factors on enrollment status found significant positive associations for three out of four factors: trust in the community (horizontal) and trust in the HCP and NHIS (both vertical) (Table 3). No association found between enrollment status and communal action.	Consistent for three out of four factors.
Traditional family support structures gradually fading due to social development and modernization. Group structures are increasingly important. Existence of many groups.	Regression of social capital factor on enrollment status ‘Previously enrolled’ showed a positive association for ‘trust in the NHIS’. All other factors showed no significant correlation.	Whereas perspectives revealed people engage in groups and social action, this shows no significant positive associations with active membership. Possible explanations: social action on solidarity/reciprocal support in the community does not focus on health; lack of interest in health issues.
General NHIS awareness among communities. Value health insurance concept (reduced financial risks when ill).		Communities value insurance concept, despite the fact they are not active members. This could explain the positive association between previous enrolled and trust in the insurance. Qualitative findings revealed clients’ reluctance to subscribe due to services not meeting their expectations. This reduces their trust in the services.
Trust in NHIS services dependent on reliable quality NHIS/DHIS and healthcare providers. Trust relations influenced by experienced challenges, i.e., attitude of staff, reliability of information and benefits package, unfavorable treatment for insured, insufficient monitoring.