Advice on comparing two independent samples of circular data in biology

Abstract

Many biological variables are recorded on a circular scale and therefore need different statistical treatment. A common question that is asked of such circular data involves comparison between two groups: Are the populations from which the two samples are drawn differently distributed around the circle? We compared 18 tests for such situations (by simulation) in terms of both abilities to control Type-I error rate near the nominal value, and statistical power. We found that only eight tests offered good control of Type-I error in all our simulated situations. Of these eight, we were able to identify the Watson’s U2 test and a MANOVA approach, based on trigonometric functions of the data, as offering the best power in the overwhelming majority of our test circumstances. There was often little to choose between these tests in terms of power, and no situation where either of the remaining six tests offered substantially better power than either of these. Hence, we recommend the routine use of either Watson’s U2 test or MANOVA approach when comparing two samples of circular data.

Distinct roles of dentate gyrus and medial entorhinal cortex inputs for phase precession and temporal correlations in the hippocampal CA3 area

“…As suggested by Landler et al, (2021), we used a one-factor MANOVA test to test for differences in the mean of circular data (e.g., theta phase). Each phase θ was treated as a single observation with two response variables cos(θ) and sin(θ), with the experimental group (control or lesion) as the sole factor.…”

The importance of time of day for magnetic body alignment in songbirds

“…We anticipate that the machine learning approach outlined in this study could be easily implemented in other animal systems, and could be implemented to create a solid framework to study body alignment and the temporal pattern of this behaviour in different settings. Moreover, although large sample size is always desirable, in cases where it is difficult or not possible to use traditional circular statistics (Batschelet 1981 ; Landler et al 2018 , 2021 ), as for our migratory birds which passage is limited in a restricted seasonal migratory period, modern modelling techniques can generate robust predictions by leveraging repeated measures of a relatively small sample size (Pewsey and García-Portugués 2021 ). Furthermore, circular mixed-effect models will allow to make inferences in experimental settings where contrasting factors are at play, like in cue-conflict experiments, because they are also capable to handle multiple covariate predictors (Nuñez-Antonio and Gutiérrez-Peña 2014 ; Pewsey and García-Portugués 2021 ).…”

Spatiotemporal organization of human sensorimotor beta burst activity

“…Test statistic is reported as value of KS (i.e., mean and range across subjects). Two circular distributions (e.g., propagation direction pre- and post-movement) were compared using two-sample Watson’s U 2 test (Landler et al, 2021) on the single subject level. Test statistic is reported as value of U 2 test (i.e., mean and range across subjects).…”