Statistics
Statistical analysis of water maze data
The original advice, short and to the point.
The original site's advice on statistics was brief and sensible, and it's reproduced here unchanged. Repeated-measures designs fit learning trials because the same animal is tested again and again.
Statistical Analysis.
When looking for statistical differences in latencies, pathlengths and swim speeds one uses repeated-measures analysis of variance (ANOVA). Assuming a significant overall main effect of treatment, genotype etc. one can employ post hoc comparisons e.g. Dunnett's test. For probe data one-way ANOVA is sufficient.
Original text: watermaze.org Part I, General Information. From original material by Richard Baker. Reproduced unchanged. Copyright Richard Baker© 2013. All Rights Reserved.
Questions people ask
Which statistics software can I use with the exported results?
Any package that reads CSV: SPSS, the free alternative PSPP, or Excel; MATLAB is also supported.
More detail
You can use any statistics package that imports comma-separated text, such as SPSS, PSPP, R, MATLAB or Excel. Exports are comma-separated (.cst or .csv), one row per trial, so they import directly. In Excel use Data > From Text/CSV (or Get External Data), delimiter Comma.
Going deeper
Because each row is one trial, repeated-measures designs need the data arranged by subject and day: use the subject and descriptor columns to reshape it, and check your regional number format uses a full stop as the decimal separator so values import correctly.
Spotted a mistake, or know a better source? Tell us and we'll fix it and say so on our corrections page.