Related Experiment Video
Updated: Feb 7, 2026

Novel Object Recognition Test for the Investigation of Learning and Memory in Mice
Published on: August 30, 2017
The Discrimination Ratio derived from Novel Object Recognition tasks as a Measure of Recognition Memory Sensitivity,
Magali H Sivakumaran1, Andrew K Mackenzie2, Imogen R Callan1
1University of St Andrews, School of Psychology and Neuroscience, St Andrews, KY16 8JP, United Kingdom.
Abstract:
Translational recognition memory research makes frequent use of the Novel Object Recognition (NOR) paradigm in which animals are simultaneously presented with one new and one old object. The preferential exploration of the new as compared to the old object produces a metric, the Discrimination Ratio (DR), assumed to represent recognition memory sensitivity. Human recognition memory studies typically assess performance using signal detection theory derived measures; sensitivity (d') and bias (c). How DR relates to d' and c and whether they measure the same underlying cognitive mechanism is, however, unknown. We investigated the correspondence between DR (eye-tracking-determined), d' and c in a sample of 37 humans. We used dwell times during a visual paired comparison task (analogous to the NOR) to determine DR, and a separate single item recognition task to derive estimates of response sensitivity and bias. DR was found to be significantly positively correlated to sensitivity but not bias. Our findings confirm that DR corresponds to d', the primary measure of recognition memory sensitivity in humans, and appears not to reflect bias. These findings are the first of their kind to suggest that animal researchers should be confident in interpreting the DR as an analogue of recognition memory sensitivity.
Related Concept Videos
Stereotypes, Prejudice, and Discrimination
Measurement: Derived Units
Hindsight Biases
Confirmation Biases
Ratio Level of Measurement
A set of data measured using the ratio scale takes care of the ratio problem and provides complete information. Ratio scale data are like interval scale data, except they have a zero point and ratios can be calculated....
Bias
In statistics, a sampling bias is created when a sample is collected from a population, and some members of the population are not as likely to be chosen as others (remember, each member...

