Methods for measuring mammalian personalities: In which animals and how accurately can we quantify it? [Metody měření personality savců: U kterých zvířat a jak přesně ji dokážeme měřit?]
Barbora Žampachová, Eva Landová, Daniel Frynta
The study of personality, and individual differences in behaviour has experienced a steady rise in popularity in the past years. In this review and meta-analysis, we aim to introduce the concept of personality and related phenomena. A behavioural trait should meet two basic conditions to be considered a personality trait – it should be consistent (1) in time and (2) across contexts. In mammals, the two most common orders in personality studies are primates and rodents. We therefore introduce different approaches to personality testing in these two orders. Primate personality studies are based on psychology studies and often rely on the observer’s ratings. Rodent personality studies originate in the studies of physiology and use an experimental approach. We present a more detailed overview of methodological issues of repeatability as a statistical tool for measuring consistency across time. The classic methods of computing repeatability do not consider habituation and other trends which may become confounding factors and lead to underestimation of repeatability. We also discuss consistency across contexts and different understandings of the context definition. We illustrate the variability of personality studies in mammals with a meta-analysis of repeatability estimates. We found that repeatability of behaviour depends on the methodology of behavioural testing and statistical analyses used, but also the number of test repetitions and differences between the focal behaviours. Repeatability decreased with more repetitions and the tests of aggressiveness and exploratory behaviour yielded lower repeatability estimates than the tests of activity.