2 Designing research questions

You may already know the area or problem you wish to research. However, rendering it researchable is not always straightforward. Critical to the success of the research project is generating research questions that have boundaries and are manageable. The following discussion explores how an interest in the role of institutions was conceptualised to render it researchable.

As you should recall, the concept of ‘institutions’ has a dual meaning:

  • as ‘complexes of norms and behaviours that serve a collective purpose’ (de Janvry, 1993); and
  • as organizations, when they are established (‘institutionalised’) and embody established norms and roles.

All innovation takes place in an institutional setting and must to some extent imply a deliberate attempt to change institutions: either the promotion of change within an institution or the establishment of a new institution. For example, the spread of an innovative use of IT will require organizational change as well as changes to the norms and roles within organizations. A significant change in, say, environmental policy will at least require change in the roles played by existing agencies if not more sweeping organizational changes and changes in underlying values. Hence an important aspect of innovation research is research about deliberate attempts at change in institutions.

Many questions about how innovation can be promoted (or ‘institutionalised’) can be stated in the following general form:

  • How can change X be implemented successfully in circumstances Y?

For example:

  • How can IT be used innovatively in primary schools?
  • How can skills and capacities be built in developing countries?
  • How can innovation be promoted in small companies?
  • How can local authorities promote sustainable development?
  • How can NGOs influence environmental policy?

These questions, stated in this way, are policy questions rather than research questions. In order to turn them into questions which are researchable one has to ask more about the ‘circumstances’. For example:

  • Who (what organization or agency) is trying to implement this change?
  • What criteria will be used to judge what is ‘successful’? What values underlie these criteria?
  • What institutions (established organizations, networks, norms) is the change to be effected within?

One problem about researching ‘How can…?’ questions like those above is that one cannot find out about things which have not yet happened. One can only speculate. So it is important to investigate cases where innovations have occurred or at least where there have been clear-cut attempts at innovations. Yet the likelihood, particularly in the early days of an attempted institutional change, is that there are few clear successes and perhaps many more examples of failure or no attempted change at all. This points to the use of case study as the most appropriate method, focusing on whatever successful cases exist, together perhaps with one or two cases where definite attempts failed in a way which promises to be illuminating.

So the general ‘How can…?’ question can be turned into a researchable question of the following general form:

  • In those cases where attempts at X have been successful, how were they done and why have they been successful?

Turning such policy questions into researchable questions will usually entail bringing in some specificity about the ‘circumstances’. For example, are we interested in the efforts of particular implementing agencies, or in attempts at innovation in a particular region or sector?

Activity 33

Think of an innovation, the possible implementation of which is close to your own research interests. Try going through the following sequence:

Question 25

Write a policy question of the general ‘How can…?’ form above which relates to the innovation you are interested in.

Question 26

Answer (briefly) the above questions about the ‘circumstances’.

Question 27

What problems do you see in trying to research this policy question directly? How do they relate to the brief discussion above?

Question 28

Now turn your policy question into a researchable question of the form suggested above.


Not everyone’s research interests can be turned into a question of exactly the same form. But I hope you were able to see that there are common kinds of problems in researching innovation or change directly which may be overcome by concentrating on cases of success. Indeed, even with a research question which prioritises cases of success like this, one can still include other cases in one’s investigations. The general form of research question put forward here is only one way of investigating innovation in institutions – although I would suggest it is an important one.

This example demonstrates how to move from an open and broad question, to a series of more discrete researchable questions. Armed with your research questions you can then identify what your knowledge needs are in order to answer them. Each research question may need a particular research method to arrive at the data evidence you require. Thus, you will have to consider in some detail your overall research design, informed by what is manageable from the perspective of your resources and time.

1.6 Managing research

3 Research design