Views from senior Australian cancer researchers on evaluating the impact of their research: results from a brief survey

Background: The interest and activity in measuring and reporting the impact of publicly funded health and medical research has grown rapidly in recent years. Research evaluation typically relies on researchers for much of the information for an impact assessment. However, the acceptability and feasibility of this activity among health researchers is unknown. The aim of this study was to understand the role and opinions of cancer researchers in the growing area of impact evaluation activity, to inform the logistics of a sustainable program of impact evaluation. Methods: A brief anonymous online survey was administered to 95 current and past grant recipients funded through the external grants program at Cancer Council New South Wales. Eleven survey statements were constructed with Likert responses and supplemented with two open-ended questions. The statements covered the conceptual, attitudinal and practical aspects of impact evaluation. The survey targeted researchers from the full spectrum of cancer control research classifications. Descriptive analyses obtained response frequencies and percentages. Results: Forty-five cancer researchers completed the survey (response rate 47%) and 77% were Associate Professors or Professors. Responses were polarised for questions relating to engaging with research end-users, perceived time-pressure to collate data, and pressure to produce research outputs. Some researchers emphasised that quality was an important goal over quantity and warned that collecting impact data created incentives and disincentives for researchers. Conclusion: There was mixed support and acceptance among senior cancer researchers in Australia on their perceived role and engagement with research impact activities. Sole reliance on researchers for collating and reporting impact data may be problematic. Requesting information from researchers could be minimised and confined to final reports and possible verification of externally-led evaluations.