Table 1 A non-exhaustive list of types of questions and research designs
From: Persistent misunderstandings about evidence-based (sorry: informed!) policy-making
Research questions | Research designs to be prioritized |
|---|---|
A. Questions about interventions | |
Is the intervention effective and/or harmful? | Intervention under the control of the researchers/evaluators |
(usually for short-term assessment) | |
∘ Randomized controlled experiments | |
∘ Prospective non-randomized controlled experiments | |
(no pure randomization procedure) | |
∘ Controlled-before-and-after study | |
Intervention NOT under the control of the researchers/evaluators | |
and/or long term effectiveness data are needed (usually for long term assessment) | |
∘ Interrupted time series | |
∘ Prospective cohort studies | |
∘ Case-control studies | |
∘ Cross-sectional studies | |
Why is the intervention ineffective or less effective than expected? | ∘ In depth mixed-method case studies |
∘ In depth qualitative case studies | |
∘ Cross-sectional studies | |
What are the stakeholders’ opinions and views about the intervention? | When targeted stakeholders are numerous and from different jurisdictions, organizations or organizational units AND high external validity is targeted |
∘ Cross-sectional opinion surveys | |
When targeted stakeholders are few and located in few settings | |
∘ In-depth mixed-methods studies including an opinion survey and | |
qualitative interviews | |
∘ In-depth qualitative studies with interviews only | |
Is the intervention cost-effective? | ∘ Formal economic evaluations |
What are the characteristics of the intervention? | ∘ Any type of studies that include an objective and a factual description of the intervention. |
Is the intervention transferable to another context? | ∘ Specific components of the study (e.g. population, environment, etc.) and the comparability of contexts need to be systematically assessed. |
B. Questions about non-interventional correlates of specified outcome(s) | |
Are correlates associated with the outcome(s)? | ∘ Prospective cohort studies |
∘ Retrospective cohort studies | |
∘ Case-control studies | |
∘ Cross-sectional studies | |
To what extent are correlates associated with the outcome(s)? | ∘ Prospective cohort studies |
∘ Retrospective cohort studies | |
∘ Case-control studies | |
∘ Cross-sectional studies | |
Why correlates might be (or might not be) associated with the outcome(s)? | ∘ Studies reporting the results of an empirical test of two (or more) |
competing theoretical explanations | |
∘ Studies reporting the results of an empirical test of solely one | |
theoretical explanation | |
∘ Narrative, opinion and textual papers (including formal theorizing | |
or not) reporting untested theoretical explanation(s) | |
C. Questions about specified outcome(s) | |
What is the prevalence of the outcome? | ∘ Cross-sectional studies |
What is the incidence of the outcome? | ∘ Prospective cohort studies |
∘ Retrospective cohort studies | |
∘ Panel studies | |
What is the economic burden of the outcome? | ∘ Any studies reporting validated estimates of the economic burden of |
the outcome for the targeted time frame, population and settings | |
How important is the outcome for stakeholders? | If targeted stakeholders are numerous and from different jurisdictions, organizations or organizational units AND high external validity is targeted |
∘ Cross-sectional opinion surveys | |
If targeted stakeholders are few and located in few settings | |
∘ In depth mixed-methods studies including an opinion survey and | |
qualitative interviews | |
∘ In depth qualitative studies with interviews only | |