Poster #3 - Children’s Pattern Predictions Show Cognitive Encoding Following Sequence Learning

• In Event: 3-053 - Poster Session 10
  In Poster Session: PS 10 Section - Attention, Learning, Memory

Integrative Statement

The longstanding debate on the role of implicit and explicit memory as separate systems (Squire, 2004) was challenged by Representational Redescription (Karmiloff-Smith, 1992) with evidence from children, suggesting a more dynamic approach than a dual memory system. Instead of systems distinguished by either limited awareness, or conscious retrieval of information after encoding, the framework outlines four levels of representations. Only the final level contains conscious representations available to verbal report, regularly used to test explicit memory in adults. This study set out to test the levels in the framework before verbal access is achieved. This is important, as children are often unable to explain or verbalise learning.

55 children between 5-11 years (32 females) were tested on four different versions of a touchscreen Serial Reaction Time task. Single- and four-image task versions featured either familiar animals or novel shapes (see Figure 1) in a counterbalanced within-subjects design. All tasks presented a multi-block design with a repeating 8-step deterministic spatiotemporal sequence mixed with random sequences. Stimulus appeared in a 2x2 grid and children were asked to tap an image as soon as it appeared on the screen. In the four-image versions, deterministic stimuli was bound to location. Afterwards, children were asked to explicitly indicate whether they recognised a pattern and predict ‘what came next and where’ using cards.

Response time analyses showed that there was a significant main effect of trial type (F[1,52] = 85.189, p < 0.001) with faster performance on deterministic over random trials for all ages and tasks. Also, a significant interaction between trial type and block (F[3, 156] = 5.167, p = 0.002) was driven by faster response times in deterministic trials over blocks (see Figure 2). A main effect of the number of stimuli in a task was found (F[1, 52] = 20.859, p < 0.001) with children tapping faster with four stimuli. There was also a significant interaction between trial type, number of stimuli and stimulus type (F[1, 52] = 4.814, p = 0.033) as children performed better on deterministic trials with four stimuli compared with one for the shapes, but not the animals. However when indicating recognition of the pattern, younger children, 5-6 years old, were significantly above chance (p = 0.041) choosing the animal pattern of the deterministic sequence tapped. Similarly for pattern prediction, a main effect of stimulus type approached significance (F[1, 52] = 3.409, p = 0.068) as children were better at pattern prediction of animals tapped rather than shapes.

These results suggest differential learning dependent on item presented. An advantage of four stimuli over one in learning the deterministic pattern supports a multiple cue theory that more information (i.e., location bound to identity) enhances learning. As only pattern prediction and preference seem learned more effectively with animals, this suggests a role for familiarity only in visual recall after motor skill or spatial learning of a pattern. The dissociation between reaction times and pattern prediction could support a more cognitive redescription, in line with the Karmiloff-Smith framework.

Authors

• Suzanne Pahlman, Centre for Brain and Cognitive Development, Birkbeck, University of London
  Presenting Author

• Natasha Z. Kirkham, Birkbeck, University of London
  Non-Presenting Author

• Annette Karmiloff-Smith, Birkbeck, University of London
  Non-Presenting Author