Poster #5 - The development of proactive interference resolution in visual working memory

• In Event: Poster Session 12

Abstract

Introduction: Proactive interference (PI) occurs when memories that were learned previously compete with currently relevant information (Keppel & Underwood, 1962; Kane & Engle, 2000). Despite extensive literature investigating the effect in adults, little work has been done in young children (Kail, 2002). In two preregistered experiments, using a Delayed Match Retrieval paradigm (Kaldy et al., 2016), we found that 3-year-old toddlers were highly sensitive to the effects of PI (Hamilton et al., submitted). Here, we used the same paradigm with a wider age range to investigate how PI resolution skills develop from early to middle childhood (preregistration: https://osf.io/dxf49).
Participants: Data has been collected from 133 children (88% of the pre-registered total sample size to date) between the ages of 2.5 and 7.5 (M = 4.9 years, SD = 1.13) at the Children’s Museum of New Hampshire. The younger group (2.5-5.5 years, N = 69) was asked to memorize two cards (see Figure 1), while the older group (4.5-7.5 years, N = 64) was to memorize three to avoid ceiling effects in performance. Half of the children between 4.5 and 5.5 years were tested with the 2-card version, and the other half were tested with the 3-card version. (This overlap allows us to further test whether set size has an effect on PI.)
Stimuli and Procedure: Participants were presented with a set of cards (on a touch-screen tablet), first face-up. Cards depicted unfamiliar objects from the NOUN database (Horst & Hout, 2016). After this initial encoding period, the cards were turned face-down, and a target card was revealed that matched one of the previously seen cards. The participant was asked to touch the (now, face-down) card that matched the target. Proactive interference (PI) was induced by repeating the same set of cards across trials. A session consisted of two blocked conditions (6 trials each, with order counterbalanced across participants): (1) PI condition: the same cards were repeated in each trial (with, of course, location randomized) and (2) No_PI condition: a new, unique set of cards was presented in each trial.
Results and Discussion: We calculated the PI effect by taking the difference in performance (mean proportion correct) in the PI vs. No_PI conditions. Our preliminary results suggest that the effect of interference accumulated over trials. A trial-by-trial analysis showed that the PI effect emerged in the second half of the trial block (t(131) = 2.37, p = 0.019, see Figure 2). The PI effect decreased with age in the 2-card task, and it was not significantly changing with age in the 3-card task. That is, the youngest children (3-year-olds) had the largest PI effect (similar to the results of Hamilton et al., submitted) (Set size did not have an effect on the ability to resolve PI in the 4.5-5.5-year-old subsample that had representation in both the 2- and 3-card versions). Thus, working memory in young children is limited by their difficulty with inhibiting interference from recent, but currently irrelevant, information.

Authors

• Mollie Dalton Hamilton, University of Massachusetts Boston
  Presenting Author

• Erik Blaser, University of Massachusetts Boston

• Zsuzsa Kaldy, University of Massachusetts Boston