2017/03/23 by Mike Thelwall, Ruth Fairclough
Computer Science · Decision Sciences · Mathematics · #Academic Publishing and Open Access #Artificial intelligence #Bootstrapping (finance) #CDF-based nonparametric confidence interval #Citation #Citation impact #Computer science #Confidence interval #Contrast (vision) #Credible interval #Econometrics #Field (mathematics) #Mathematics #Research Data Management Practices #Robust confidence intervals #Sample (material) #Sample size determination #Statistics #cs.DL #scientometrics and bibliometrics research
paper · pdf · doi:10.1016/j.joi.2017.03.004
Journal of Informetrics, in press
arxiv created 2017/03/23 · arxiv updated 2017/03/24 · openalex publication_date 2017/04/07 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
When comparing the average citation impact of research groups, universities and countries, field normalisation reduces the influence of discipline and time. Confidence intervals for these indicators can help with attempts to infer whether differences between sets of publications are due to chance factors. Although both bootstrapping and formulae have been proposed for these, their accuracy is unknown. In response, this article uses simulated data to systematically compare the accuracy of confidence limits in the simplest possible case, a single field and year. The results suggest that the MNLCS (Mean Normalised Log-transformed Citation Score) confidence interval formula is conservative for large groups but almost always safe, whereas bootstrap MNLCS confidence intervals tend to be accurate but can be unsafe for smaller world or group sample sizes. In contrast, bootstrap MNCS (Mean Normalised Citation Score) confidence intervals can be very unsafe, although their accuracy increases with sample sizes.