This paper explores a 7-stage cluster methodology as a process to identify appropriate indicators for evaluation of individual researchers at a disciplinary and seniority level. Publication and citation data for 741 researchers from 4 disciplines was collected in Web of Science. Forty-four indicators of individual researcher performance were computed using the data. The clustering solution was supported by continued reference to the researcher’s curriculum vitae, an effect analysis and a risk analysis. Disciplinary appropriate indicators were identified and used to divide the researchers into four groups; low, middle, high and extremely high performers. Seniority-specific indicators were not identified. The practical importance of the recommended disciplinary appropriate indicators is concerning. Our study revealed several critical concerns that should be investigated in the application of statistics in research evaluation.
The strength of the 7-stage cluster methodology is that it makes clear that in the evaluation of individual researchers, statistics cannot stand alone. The methodology is reliant on contextual information to verify the bibliometric values and cluster solution. It is important to do studies that investigate the usefulness of statistical evaluation methodologies to help us as a community learn more about the appropriateness of particular bibliometric indicators in the analysis of different researcher profiles.