Poster #9 - Memory Interference and Generalization: Current and Cumulative Bilingual Language Exposure Provides Insight into Memory Changes

• In Event: 3-106 - Poster Session 11
  In Poster Session: PS 11 Section - Attention, Learning, Memory

Integrative Statement

Introduction of novel information can result in retroactive interference as early as 3 months of age (Rossi-George & Rovee-Collier, 1999), and memory specificity limits toddlers’ ability to generalize knowledge to novel situations. However, bilinguals demonstrate memory generalization earlier than monolinguals do (Brito & Barr, 2014, 2012). The present study examined whether there are differences in performance on generalization and interference tasks between 2-year-old monolinguals and bilinguals.
Forty-five 24-to 27-month-olds participated (current language status: 32 monolingual, 13 bilingual; cumulative language status: 34 monolingual, 11 bilingual). All toddlers’ primary language was English. There were two sets of stimuli (animals: rabbit, monkey, panda; rattles: A, B, C) and three target actions for each set of stimuli. Toddlers saw target actions demonstrated on two stimulus sets (e.g., rabbit and rattle A) and then were tested 30 minutes later on 4 stimuli: a different animal (e.g., monkey), a different rattle (e.g., rattle B), which constitutes generalization, and the same animal (e.g., rabbit), and the same rattle (e.g., rattle A), which constitutes interference. Imitation scores were calculated for each stimulus (max = 3 target actions), as well as a composite score (max = 6) for generalization and interference. The Language Exposure Assessment Tool (LEAT; DeAnda et al., 2016) was used to measure bilingual status both currently (at 24 months) and cumulatively (over the past 24 months).
Two Poisson models were conducted on the full sample to predict scores on generalization based on bilingual status (one model for cumulative and one model for current), gender, median income, and percentile of English vocabulary scores on the MacArthur-Bates Communicative Development Inventory (MCDI; Fenson et al., 2000). There were no significant results in either model. Two Poisson models were conducted to predict scores on interference based on the same variables as above. Current bilinguals (p = .02) and cumulative bilinguals (p = .052) performed better than monolinguals did. Additional analysis matching monolingual and bilingual children on their English MCDI scores, the median income in the family’s zipcode, and the child’s gender yielded a similar pattern of results. In the matched sample, each participant’s cumulative bilingual status equaled their current bilingual status. There were no significant differences between bilinguals and monolinguals on generalization, but there was a trend for bilinguals to perform better than monolinguals on interference, p = .071.
In sum, there were no differences on generalization, which does not replicate prior findings but may be due to some changes in stimuli which we are currently investigating. However, there were differences on interference, with bilinguals showing less interference than monolinguals. By using the LEAT, the current findings extend prior findings by indexing both cumulative bilingual status as well as the more commonly used current bilingual status, and indicate that results differ depending on how bilingual status is indexed. Most studies measure current but not cumulative bilingual status and both metrics were assessed in conjunction with performance on tasks.

Authors

• Joscelin Rocha Hidalgo, Georgetown University
  Presenting Author

• Sylvia N Rusnak, Georgetown University
  Non-Presenting Author

• Olivia Blanchfield, Georgetown University
  Non-Presenting Author

• Alison K Schug, Georgetown University
  Non-Presenting Author

• Elizabeth Holland, Colby College
  Non-Presenting Author

• Eva Nunez, Georgetown University
  Non-Presenting Author

• Sharanya Suresh, Cornell University
  Non-Presenting Author

• Rachel Barr, Georgetown University
  Non-Presenting Author