Abstract
When an object that is moving along a particular path vanishes, observers' memory for the object's final position is biased in the direction of continuing motion (Freyd & Finke, 1984). This effect is seen even with static stimuli in which motion is simply implied (Freyd, 1983). This memory bias, referred to as the representational momentum (RM) effect, is seen in both children and adults (Futterweit & Beilin, 1994; Hubbard, Matzenbacher, & Davis 1999). Several recent studies (e.g., Senior et al., 2002) suggest that images depicting implied motion activate brain regions involved in actual motion processing. Given this, we might expect to see that children with motion-processing deficits would show an atypical RM effect. Previous research in our laboratory has shown that children born prematurely at very low birth weight (VLBW [[lt]] 1500 g) are at risk for impairments in both low-level and high-level motion processing (MacKay et al., 2005; Jakobson et al., 2006). In the present study, we compared the RM effect in 5–9 year old, VLBW children to that seen in an age-matched sample of full-term controls. Full-term children showed a robust RM effect, and the strength of this effect was negatively correlated with their global motion coherence thresholds (r = −.55); in other words, control children who were more proficient at global motion perception showed a larger memory bias. VLBW children did not show an RM effect, and tended to have higher motion coherence thresholds than controls, overall. There was, moreover, no relationship between performance on the RM task and motion coherence thresholds (r = −.06) in this group. These results are consistent with research suggesting overlap in the neural substrates supporting the processing of implied and actual motion, and extend our earlier work demonstrating that VLBW children are at risk for problems associated with poor motion processing.
This research was supported by an NSERC grant to LJ