Rodent Cognition Model

Rodent cognition models convert an unobservable process into a number an experimenter can record. The Morris water maze, introduced by Richard Morris in 1981, requires an animal to locate a hidden platform in opaque water using distal cues, scored by escape latency across training days and by time in the target quadrant on a probe trial. Novel object recognition exploits an innate preference for novelty and yields a discrimination index. The Y-maze scores spontaneous alternation as a working memory proxy, the Barnes maze substitutes a dry aversive surface, and fear conditioning measures freezing to a cue or context.

Each task measures performance, which is not the same as memory. Water maze latency is sensitive to swim speed, buoyancy, thigmotaxis and the stress of forced swimming; a sedating or anxiolytic compound can shift the score without touching hippocampal function. Novel object recognition depends on handling, illumination, olfactory residue and exploration time, and its effect sizes are small relative to the group sizes typically used, often eight to twelve animals per arm. Strain matters: some inbred lines are visually impaired and perform poorly for reasons unrelated to cognition.

Used properly these assays screen mechanisms and establish dose-response where invasive confirmation is possible. What they cannot do is model the clinical syndromes their results are usually mapped onto, because the human endpoints of interest are episodic memory, executive function and everyday capability, none of which have rodent equivalents.

The error encountered most often is a single behavioural figure presented as evidence of memory enhancement in humans. One maze, one timepoint, one sex, no blinding of the observer, and no correction for the several endpoints each apparatus generates is a hypothesis, not a cognitive claim.