Platform crossings are widely used in Morris water maze analysis as an indicator of spatial memory in probe trials. They are also useful in reversal learning, quantifying perseveration – searching the previously learned location when the platform has been moved.

Platform crossing counts provide a more localized measure than quadrant preference, and are useful regardless of the platform position (not just when it is in the center of a quadrant); nonetheless it is important to take account of the limitations of this measure.

Firstly, the count of exact platform crossings may underestimate performance when search is concentrated near, but not precisely at, the target – it does not capture near misses or close approaches, where the subject shows imprecise but spatially directed search. For this reason the HVS Image system counts both the number of crossings of the exact location the platform was in, and the number of close encounters.

Both exact crossings and close encounter measures can be given for each calibrated platform position, allowing quantification of behavior relating to a previous platform position (e.g. in reversal learning) or another position of interest (e.g. in dual strategy trials), as well as behavior relating to the actual platform or learned position. For clarity, the number of platform crossings for the actual learned position is set out separately from the set of counts for each calibrated position.

A second limitation is that platform crossings counted over the entire duration of a probe trial may suggest stronger spatial memory in subjects who continue to search the learned location (and cross the platform position many times while searching the area), than in subjects who have both precise spatial memory and cognitive flexibility – the latter determining quickly that the platform is not in the learned location and moving on to search elsewhere, with fewer platform crossings.

It is therefore important to use additional analyses such as heading angle (to quantify spatial intent at the start of the trial), time and path to the first close encounter, and time slice analysis to allow you to distinguish between behaviors early and later in the trial.

In addition, HVS Image’s automated behavior identification gives you a summary of behaviors, including identifying direct path to the target area, focussed search of the target area and/or target scanning, even if the subject subsequently engages in behavior that results in relatively low platform crossings over the full duration of the trial, such as searching elsewhere or reverting to a random path or thigmotaxis.

See citation and technical note.

Explore all the Morris water maze measures and analyses.