The effect of experimenter experience on behavioural measures in zebrafish: Comparison of observation-based and video-tracking methods

The zebrafish is a frequently employed model organism in neuroscience. Behavioural analysis is often a critical component of neuroscience studies. Issues with replicability and reproducibility of experimental results have been in the focus in several fields of biology. In behavioural studies, one of the possible reasons for such issues is experimenter experience, i.e., the ability of the experimenter to properly quantify behaviour of their experimental animals. In this proof-of-concept study, we investigate how different levels and types of experimenter experience may influence results when the experimenter employs automated video-tracking or observation-based software to quantify behaviour of zebrafish. For example, we investigated whether experience with the particular software application, or with maintenance of zebrafish, or with zebrafish behaviour makes a difference. To answer these questions, multiple students, with different levels and types of prior experience, analyzed the video-recordings of the same 18 zebrafish. Subsequently, we analyzed the correlation coefficients between distinct experience types/levels and a variety of behavioural measures. We found that experience in zebrafish maintenance or behaviour did not correlate significantly with any behavioural measures quantified using video-tracking, but level of experience with the video-tracking software did with some measures. Conversely, experience with zebrafish maintenance and with zebrafish behaviour significantly correlated with some behaviours measured using the observation-based method, whereas level of experience with the software employed for this method did not. We conclude that the effect of the level and type of experience the experimenter has may affect results in a quantification method-dependent and behaviour-specific manner in the analysis of zebrafish behaviour.