There are many challenges in undertaking work to assess the impact of research on policy. The study of research impact and research-policy linkages is simply not amenable to a single conventional means of analysis. Nevertheless, any work in this area must evaluate the evidence in as rigorous a manner as possible and there is a growing body of experience that indicates the value of using context-specific, triangulated combinations of qualitative and quantitative methods. Our intention here is to highlight a number of challenges and to provide some suggestions that might enable any would-be researchers to address these challenges.

There are a number of specific challenges that researchers in this area will need to confront. These include:

• It is difficult to quantify and attribute policy impact. Therefore, there is a need to work from both the research end (to assess policy impact of specific bits of research) and policy end (to assess the role of research in clear policy shifts).
• Situations tend to be very specific and it is difficult to find middle ground between lessons which are so context specific as to be inapplicable elsewhere, or too generalised to be practically useful. The implication is that we need more detailed, in-depth studies of specific events and broader studies looking across several contexts.
• There are too often many variables to draw clear conclusions. These might include issues such as: contested evidence, different political contexts and multiple networks and communication. One solution is to find ways of reducing the number of variables, for example by focusing on specific relatively uncontested evidence (how to stop the spread of HIV/AIDS) and compare policy adoption in different countries. Another way would be to evaluate the impact of evidence from different research projects in one sector in one country. A third would be to compare the effectiveness of different policy research institutes and look for the characteristics that were influential.
• Much of the evidence is qualitative and subjective. Therefore there is a need to triangulate, using several different research approaches to investigate the same issues.

There are various methodological approaches (both qualitative and quantitative) that may be used in analysing issues of research policy linkage. Some of the options – and sources for further information on each – are outlined below:

• Reviews – There have been general reviews on this issue (for example by Sutton, Nielson and Hovland et al), but there is still some scope for specific reviews.
• Case study analysis – For example, the first phase of the GDN Bridging Research and Policy project collected 50 case studies. http://www.gdnet.org/rapnet/research/studies/case_studies/Case_Studies_Index.html
• Studies of Research Impact – IFPRI has undertaken a number of studies in this area and IDRC is currently completing a systematic evaluation of the policy impact of its research programme.
• Episode studies – These refer to a study that focuses on a clear policy change and tracks back to assess what impact research had among the key issues driving the change. The could be focusing on a single episode or comparative. For further information see. ODI's RAPID Programme is using this approach in its detailed case studies.
• Interviews – Standard interview techniques are likely to form a critical component of most investigations into research-policy links.
• Surveys – It is possible to use survey techniques – see preliminary surveys of researchers and policymakers in the GDN project. http://www.gdnet.org/rapnet/research/studies/surveys/
• Rates of return analysis – some studies have attempted to measure the rates of return of research.
• Write shops of policymakers and communicators – This technique provides a way to tap the knowledge and experience of relevant policymakers, practitioners and communicators.
• Sector Studies – The work of IFPRI provides a rare example of a sector study (on food policy) although from the perspective of a single organization.
• Comparison of countries / sectors – At present there has been very little systematic work on Bridging Research and Policy within or across countries. These could include a combination of approaches.

Given the difficulties of undertaking research in this area, it is often useful to combine several approaches and use 'triangulation' to verify the results.