Abstract
This study examined 6-month-old infants' abilities to use the visual information provided by simulated self-movement through the world, and movement of an object through the world, for spatial orientation. Infants were habituated to a visual display in which they saw a toy hidden, followed by either rotation of the point of observation through the world (simulated self-movement) or movement of the object itself through the world (object movement). Following habituation, infants saw test displays in which the hidden toy reappeared at the correct or incorrect location, relative to the earlier movements. Infants habituated to simulated self-movement looked longer at the recovery of the toy from an incorrect, relative to correct location. In contrast, infants habituated to object movement showed no differential looking to either correct or incorrect test displays. These findings are discussed within a theoretical framework of spatial orientation emphasizing the availability and use of spatial information.
Notes
One worrisome issue involving the looking times toward the test events is that because the minimum look criterion built into the habituation procedure was shorter than the length of the event sequence itself, infants might have look durations shorter than it would take for a complete repetition of the test event. In response to this concern, the data for the test trials were examined for all fixations less than 4.0 sec, which was roughly the amount of time required to see the object reappear out of the cup in the two dishabituation events. For the object movement condition, 25% and 30% of the trials fell below this critical time for the correct and incorrect trials, respectively; for the simulated self-movement condition, 32.5% and 10% of trials fell below this critical time. Although there do appear to be a substantial proportion of trials falling below this critical value, there are reasons not to be overly concerned by this fact. First, despite there being some trials in which looking times were indeed short, infants' mean looking on the test trials was long enough to have seen the entire test sequence at least twice (between 12 sec and 16 sec); hence, they clearly did generally see the critical reappearance information on average. Second, a comparable analysis to that reported in the text of looking times to the test events, removing data falling below this critical time, produced similar results, with the Test Display × Movement condition almost reaching significance, F(1, 15) = 4.05, MSE = 180.77, p = .06. The fact that this interaction is now only marginally significant is to be expected, given that removal of looking times less than 4.0 sec trims the data of those trials showing the strongest generalization between habituation and test, and thus more selectively affects the data for the correct trials of the simulated self-movement condition than any other condition (a fact borne out by the percentage of trials falling below this critical value, as described earlier).