A New Measure of Mnemonic Discrimination Applicable to Recognition Memory Tests With Continuous Variation in Novel...

Mnemonic discrimination (MD) involves distinguishing new stimuli from highly similar memories; it is impaired in the elderly and individuals with neuropsychiatric disorders and may also probe hippocampal dentate gyrus function. Measuring MD is, therefore, highly relevant; however, the gold-standard MD test, the mnemonic similarity task (MST), is rarely used in clinical research. Thus, it would be useful to develop a novel MD index applicable to recognition memory tasks that are commonly used in clinical research. The present study develops such a measure and demonstrates its convergent validity with the gold-standard MD index from the MST.

We derived participant-level indices of MD (λ) and overall recognition memory performance (Δ) by fitting a logistic function to the relationship between stimulus interference and the probability of classifying a stimulus as novel. We then applied these novel measures to two independent MST datasets (N = 18; N = 67) and to simulated MST data. We used linear mixed-effects model to test whether (1) λ predicts the MST's MD measure, the lure discrimination index (LDI), and (2) Δ predicts the MST's overall recognition memory index (REC).

λ predicted LDI (β = 0.76, 95% CI [0.62, 0.91], p < 0.001) but not REC (β = 0.06, 95% CI [-0.03, 0.15], p = 0.197), while Δ predicted REC (β = 0.93, 95% CI [0.83, 1.02], p < 0.001) but not LDI (β = -0.06, 95% CI [-0.20, 0.09], p = 0.438). The λ and Δ indices were statistically independent, although simulations with synthetic data suggest that MD measurement may be compromised if overall recognition memory performance is impaired.

We have developed a novel measure of MD applicable to two-choice recognition memory tasks that use stimuli with continuously varying degrees of similarity. Future studies should further validate this measure using large clinical datasets that include both MD and other recognition memory tasks.