Communicating knowledge : How and why researchers publish and disseminate their findings; Supporting paper 3: Report and analysis of researcher survey

Abstract

An online survey of UK academic researchers was conducted over six weeks in March to May 2009. Over 800 useable responses were received, and the following key points emerged from the analysis: Publication and dissemination behaviour • • Peer reviewed journals are seen as the most important form of research dissemination overall, with 94% of respondents rating them as ‘very important’ and 6% as ‘quite important’. Conference presentations are rated as ‘very important’ or ‘quite important’ by 86% of respondents overall, book chapters by 83%, and monographs by 59%. There are notable disciplinary differences in the relative importance of professional journals, monographs, reports, datasets and open access repositories. • The key influences in the choice of means of dissemination are career advancement, maximising the dissemination and research assessment requirements. The relative importance of these varies by output type and discipline. • There are also differences by career stage in what influences this choice. • The means of ordering authors in collaborative outputs was very varied, and differs by discipline. Most disciplines registered high use of several methods, suggesting that there are also differences between fine-grained subjects within the broader discipline areas. Citation behaviour • The highest motivations for citing material are its known authority, and the requirement to reference an original method/theory/argument. • • There were few differences in citation behaviour by discipline, but more by career stage. Early career researchers were more likely to record high levels of influences of various kinds than were more experienced researchers. Training in citation practice was not widespread. • Early career researchers were more likely to have received guidance on citation, especially from their institution. • Guidance most often came from reviewers, and this was seen as both positive (pointing out material which had been missed) and negative (suggesting citing the reviewer and/or works from the journal). Co-authors were also an important source of citation guidance. Research Assessment • The RAE is thought to have had an impact on dissemination practice, with comments suggesting that this was broadly thought to be detrimental. • Most information on research assessment came from the institution, with less than half the respondents obtaining information from funders’ websites. This may be a contributory factor in the apparently common misconception that the only outputs acceptable to the RAE were high impact journal articles. A significant minority of respondents thought that the RAE rules excluded important research outputs. • The proposed Research Excellence Framework, as it is understood by researchers, is expected to influence both dissemination and citation behaviours. • Nearly one quarter of respondents would produce more outputs, and one third would increase their submission rate to quality outlets. Forty-two percent would make more outputs available on open access. • Nearly 40% of respondents would cite collaborators’ work more often. Although few would change the frequency with which they cited competitors’ work.