Fly-in 0.5s
Encode each 1 s
Sample 1s
Response 2s
Reward 1s
Exploring representation updating in a dynamic Visual Working Memory task in 20- and 25-month-olds
Chen Cheng, Zsuzsa Kaldy, Sangya Dhungana, Erik BlaserDevelopmental and Brain Sciences, University of Massachusetts Boston, Boston, MA
• 25-month olds could dynamically update a VWM representation in response to both a translational and shuffle movement of to-be-remembered items.
• Though overall performance showed 20-month olds failed to perform the updating tasks, they showed a (weak) learning trend, suggesting further training and practice may be beneficial; further study is needed.
• A suggestive developmental gender difference was found: girls out performed boys.
Introduction
HUMAN VISION LAB
Methods
Discussion
Acknowledgement & RefsThis project was supported by NIH #1R15HD086658.Piaget, J., & Inhelder, B. (1969).1969. Richardson, D. C., & Kirkham, N. Z. (2004). Journal of Experimental Psychology: General; Barth, J., & Call, J. (2006). Journal of Experimental Psychology: Animal Behavior Processes; Kaldy, Z., Guillory, S. B., & Blaser, E. (2016). Developmental science; Voyer, D., Postma, A., Brake, B., & Imperato-McGinley, J. (2007). Psychonomic bulletin & review. For more info, please contact ([email protected])
Dependent measure: anticipatory saccade to face-down ‘Match’ (correct) vs. ‘Non-match’ (incorrect) card based on longest look during response interval
In the real world, infants must engage with dynamic events that require the updating of information in Visual Working Memory (VWM).
Can infants hold and update information in a
dynamic VWM task?
Delayed Match Retrieval (Kaldy, Guillory, & Blaser, 2016)
Results
Exp 1: N = 21, M_age = 20.2 (18-22 mo)Exp 2: N = 21, M_age = 20.0 (18-22 mo) N = 21, M_age = 24.7 (22-26 mo)Exp 3: N = 21, M_age = 19.3 (18-22 mo) N = 21, M_age = 24.9 (22-26 mo)
• 12 trials total• Only infants who completed
3+ trials were included• First three trials were
discarded.
No movement (Exp. 1) Translational movement (Exp. 2) Shuffle movement (Exp. 3)
Non-Match Match
VWM performance
Learning effect?
Gender difference?
20-month olds could remember object-location bindings only when objects did not move (Exp1); But older (25-month) infants could track object-location bindings with translational (Exp 2) or shuffle movement (Exp 3) of cards.
• In both translational (Exp 2) and shuffle (Exp 3) movement, performance at both ages was better at the end of the task, suggesting individuals may be learning the task over time.
Translational movement: both objects move to new locations, not previously occupied, and on the same side.
Shuffle movement:one moves to a new location, the other moves to a pre-occupied location.
Collapsing two updating movements together, we found an interaction effect between age group and gender: girls performed better than boys at 25-month old, but worse at 20-month old (p < 0.005).
P2-E-110
Translational (Exp. 2)
0.5
Prop
ortio
n co
rrec
t (ba
sed
on
long
er lo
ok)
0.6
0.7
0.4
0.3
0.6
20-mo 20-mo 25-mo 20-mo 25-mo
Shuffle (Exp. 3)
No movement (Exp. 1)
0.8
0.7
0.6
0.5
0.4
0.3
0.2
Trial1-3 Trial4-6 Trial7-9 Trial10-12
20-mo25-mo
While traditional VWM research h a s f o c u s e d o n s t o r a g e limitations, we aim to explore how well infants manipulate stored information, i.e. updating.
Piaget (1969) initiated the study of object transferral in the A-not-B task; later, studies found that 6-month olds were able to update information with respect to multimodal events (Richardson et al., 2004); then, at 30-month of age, children could pass invisible displacement (Barth et al., 2006).
0.8
0.7
0.6
0.5
0.4
0.3
0.2Prop
ortio
n co
rrec
t (ba
sed
on
long
er lo
ok)
Trial1-3 Trial4-6 Trial7-9 Trial10-12
Fly-in 0.5s
Encode each 1 s
Mover down 1.2s Sample1s
Response 2s
Reward 1s
Fly-in 0.5s
Encode each 1 s
Mover down 1.2s Sample1s
Response 2s
Reward 1s
Cards enter either side of the screen randomly
Participants
Translational movement (Exp. 2) Shuffle movement (Exp. 3)