The path length is the total length of the subject’s path during the trial. It’s one of the most commonly used measures in Morris water maze analysis, as it gives a simple, quantitative measure of performance.
To interpret path length correctly, it’s important to consider context, trial structure, and other behavioral indicators, as discussed below, but when used appropriately it offers insights into learning, motor function, and task engagement across a variety of paradigms, including:
- Spatial learning and memory:
Path length is widely used here, for example in hippocampal lesion studies and pharmacological interventions targeting spatial memory. It helps assess efficiency of navigation toward a hidden platform over repeated trials – reducing path lengths over learning trials indicate learning of the platform location. - Cognitive flexibility – reversal learning:
Path length helps evaluate the ability to adapt to a new platform location after learning a prior one, for example in prefrontal cortex and neurodevelopmental disorder models. Initial increases in path length after the platform is moved, followed by decreases, indicate flexible spatial adaptation. - Cue learning (visible platform):
Cue learning tests sensorimotor ability, motivation, and visual function, and is often used as a control phase in spatial memory studies. Short path lengths to a visible cue show the visual acuity, motor ability, motivation and task comprehension necessary for the study. - Sensorimotor function and neurological impairment:
Here long path lengths with poor targeting may suggest motor or balance impairments, for example due to stroke, vestibular damage, or cerebellar dysfunction. - Aging and neurodegeneration:
Path length is sensitive to age-related learning impairments, particularly in reference and working memory tasks. Longer, more erratic paths are typical in models of aging or Alzheimer’s disease, showing declines in spatial memory and mobility in aging rodents. - Stress, Anxiety, and Motivation Assessments:
Longer path lengths may suggest circling or wall-hugging associated with anxiety-like behavior in models of chronic stress and anxiety disorders, or a lack of engagement with the task, especially in early training.
It’s important to note the limitations of path length and consider which other measures are also needed for your particular study.
An obvious point is that path length is confounded by start to platform distance. In experiments where start positions vary, longer distances to the platform naturally produce longer path lengths, regardless of spatial knowledge and other competencies. Using the path efficiency ratio rather than the raw path length overcomes this.
In addition path length gives only minimal insights into strategy. A long path may for example be due to thigmotaxis, scanning the entire pool or looping close to the platform location without quite reaching it – all of which are specific, distinct behaviors. Likewise procedural strategies such as chaining may result in reducing path lengths that could be misinterpreted as spatial learning if other measures are not also used.
To understand behavior it’s important to go beyond basic measures like path length and latency. For a quick answer, use the automatic strategy classification offered in HVS Image 2025 onwards, or to see the details for yourself use a range of analyses, such as:
- Path ratio – normalizes path length by direct distance to the platform, revealing how direct the route was, regardless of start position.
- Thigmotaxis – see the percentage of the path that was spent hugging the pool wall
- Pool circling and chaining – see whether the subject found the platform by swimming in circles
- Proximity measures such as the Gallagher proximity measures and the close pass measures. In different ways these quantify how close the subject was to the platform location during the trial and are more sensitive than path length for assessing search focus, especially in probe trials.
- Percentage of path in corridor or cone – see whether the subject followed a spatially appropriate trajectory, even if its total path length was inflated by meandering or looping.
- Heading angle – captures initial spatial orientation and intention, useful in learning and probe trials.