Does variability in recognition memory scale with mean memory strength or encoding variability in the UVSD model?

Abstract

The unequal variance signal detection (UVSD) model of recognition memory assumes that the variance of old item memory strength (σ<sub>o</sub>) is typically greater than that of new items. It has been suggested that this <i>old item variance effect</i> can be explained by the encoding variability hypothesis. However, Spanton and Berry (2020) failed to find evidence for this account, suggesting that σ<sub>o</sub> may simply scale with mean memory strength (<i>d</i>) in the UVSD model. Experiments 1 and 2 examined the effects of encoding variability and strength scaling on old item variance by creating conditions in which mean memory strength and variability in item characteristics was either low or high in 2 × 2 factorial designs. In Experiment 1, overall strength determined estimates of σ<sub>o</sub>, with no effect of item characteristic variability. The same effect of overall strength was found in Experiment 2; there was also a significant effect of item characteristic variability, although this manipulation also had some effect on <i>d</i> and was therefore partially confounded. Experiment 3 similarly found a simultaneous increase in old item variance and memory strength in a design using mixed item characteristic variability conditions in a single study/test block. We conclude that old item variance increases with mean memory strength in the UVSD model, with uncertainty about the effects of encoding variability, and that future explanations of the old item variance effect should bear this in mind.

Greater Target or Lure Variability? An Exploration on the Effects of Stimulus Types and Memory Paradigms.

“…It is worth noting that, the variation in the SD parameters across serial positions was large in Experiment 2a and 2b, with target variance increased for initially studied and more recent items. This potentially adds to a growing list of variables that do affect target item variance (Spanton & Berry, 2022), with further research needed to validate such findings. One explanation for such variation is that discriminability has been found to be significantly positively correlated with target variance, in contrast to the constancy of slopes hypothesis (Spanton & Berry, 2022).…”

“…This potentially adds to a growing list of variables that do affect target item variance (Spanton & Berry, 2022), with further research needed to validate such findings. One explanation for such variation is that discriminability has been found to be significantly positively correlated with target variance, in contrast to the constancy of slopes hypothesis (Spanton & Berry, 2022). Thus, conditions of improved memory performance would be expected to show lower zROC slopes.…”

“…All participants were undergraduate students from the University of Melbourne who participated for course credits. The number of participants was selected in accordance with previous studies (e.g., Heathcote, 2003;Spanton & Berry, 2022), where 64 to 75 participants were recruited. It is important to note that in ROC studies, the number of trials per participants is likely to be more consequential than the number of participants.…”

“…This is because large numbers of observations are required to obtain stable estimates of the zROC slope, which we are attempting to do for each individual participant. In present experiment, each participant was tested on 768 trials per condition that were distributed across two 1-hour sessions, which should be sufficient when comparing to 480 trials per participant in previous studies (e.g., Spanton & Berry, 2022). The study was approved by The University of Melbourne Psychological Sciences Human Ethics Advisory Group (Ethics ID: 12033).…”

Everyday amnesia: Residual memory for high confidence misses and implications for decision models of recognition.

“…In the future, however, the predictions of other instantiations of these models that allow for the memory of individual items to change from one phase to the next could be explored. If, for example, memory weakens over time, and the variance of the old relative to the new item distribution is linked to overall strength as some have observed (e.g., Spanton & Berry, 2020, 2023), then the UVSD model predictions may be more similar to those that would be made under equal variance assumptions, and the model may therefore not mis-predict the residual memory effect for HCMs.…”