Assessment of research impact need not be all that difficult

THERE are several objective components of impact that can easily be calculated.

MEASURING research impact is back on the agenda. And, personally, I can't see why such a high degree of difficulty has been attached to the measurement of social, economic and environmental impact - it's really not that hard. One just needs to move away from the rigid concept of metrics to a more generous focus on assessment.

It also requires acknowledgment that the impact of research really only comes into play when one moves away from the outputs of the research to the point where they have a clear effect on individuals, society or firms.

The implications are that some elements of impact will be necessarily subjective, some will take a long time to be fully realised, we may need to use case studies in order to describe complex examples, end-users may have to be involved in the assessment, and we will probably have to look backwards as well as forwards in time (the British model goes back to 1994).

Nevertheless, there are several objective components of impact that can easily be calculated: patents granted; external research funding; amount of repeat funding from external sources; licences granted; royalty income received; sustainable employment and market value created from spin-out companies; and, perhaps, operating and capital benefits specifically achieved by business.

In the social sciences and humanities one can use the measure of government references made to the work in defining public policy and legislation development, or the number of books sold.

There is also value in using an assessment of the "expected impact if the research and its translation are successful". This removes the effect of encumbrances to the ultimate delivery of impact that are unrelated to the relevance or quality of the research, such as lack of investment capital, widespread public opposition, collapse of the anticipated market, or the emergence of a major competitor. These matters are beyond the control of the researchers and should not diminish the scope of the impact that would have been achieved under different circumstances of time and place.

Of course, any assessment of expected or potential impact, since it has not yet been achieved, must include a qualification for practicality, degree of difficulty, and likelihood of success. This would avoid the complete dominance of research proposals that would have massive impact, but which are very, very unlikely to succeed and may require enormous financial investment and time scales.

It has troubled me why there appears to be such a strong resistance to the assessment of impact and its formal inclusion into measures of research; in Britain, the impact component was relegated to only 20 per cent of the overall measure of research quality.

Could it be that, at some level, the inclusion of an impact parameter challenges the independence and purity of the research? Is it seen as the thin edge of the wedge of greater social accountability of research? Does it reflect the deep-seated culture in academe that payment for research is of an inherently lower quality than pure, undirected research?

I hope the new innovation dividend committee, which will provide advice to government on research impact, is given the chance to do its job through the inclusion of some of these broader parameters and in the absence of any of the entrenched research cultural constraints.

Rod Hill is pro vice-chancellor (industry engagement and commercialisation) at Monash University.