Choosing water maze software without regrets

Every package on the market will tell you how long the animal took to reach the platform. That's the easy bit. The real differences appear the moment you ask a harder question, so here's the list we'd take to any demo.

1. Can you trust the track?

• You want time-stamped positions at a steady frame rate, not just a pretty line on the screen.
• Dark animals, white animals, glare and ripples should all be handled without hand-holding.
• Pool size and position must be calibrated, so a centimetre on the report really is a centimetre.

2. Does it look past latency?

Latency and path length both improve when an animal picks up a non-spatial habit, so they flatter more studies than they should. Ask about directional measures (heading error, path efficiency, corridor and cone tests), proximity measures (close passes, Gallagher global and cumulative), and the awkward behaviours: thigmotaxis, circling, chaining and floating.

3. Will it name the strategy?

Classifying the search strategy on each trial tells you how a group solved the task. It's the cleanest way to tell a genuine spatial map from a clever procedural shortcut.

4. Will it bend to your design?

• Any start point and any platform position, not only quadrant centres.
• Reversal and dual-strategy trials with old and new positions both calibrated.
• Time slices, for animals that change their mind partway through a trial.

5. Can someone else check your work?

Raw tracks and results should export in plain formats so a colleague, or a sceptical reviewer, can re-run the analysis. That's the spirit of the FAIR principles.