The Accuracy of Confidence Intervals for Field Normalised Indicators

TL;DR

This paper evaluates the accuracy of different methods for calculating confidence intervals of field-normalized citation impact indicators, finding that some methods are conservative and others can be unsafe, especially with small samples.

Contribution

It systematically compares the accuracy of bootstrap and formula-based confidence intervals for field-normalized citation indicators using simulated data.

Findings

MNLCS confidence intervals are conservative but safe for large groups.

Bootstrap MNLCS intervals are generally accurate but can be unsafe with small samples.

Bootstrap MNCS intervals can be unsafe, improving with larger samples.

Abstract

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.