Visual bases for the perception of facial expressions : a look at some dynamic aspects

This series of experiments was designed to investigate ecological sequences of dynamic facial expressions. Several manipulations were performed to examine the visual system's sensitivity to dynamic in comparison to static facial expression displays. Experiment 1 investigated subjects' sensitivity to changes in the duration of a neutral expression embedded in a dynamic sequence. The sequences depicted a change in emotion from sadness to happiness (or vice versa). Both forward and backward presentation were utilized. The results showed sensitivity to subtle information associated with the presentation direction of a facial expression sequence. This proved to be independent of the type of presentation used, either real or induced. Experiment 2 was designed to screen for any possible artifacts (expressive noise) in the sequences and to establish a threshold of discriminability. The discriminability of the emotional expressions was found to be acute and judgment accuracy was approximately 100%. Nevertheless, this manipulation turned out to be sensitive to expressive noise present in the sequences suggesting that a sequence segment may, therefore, not accurately reflect the targeted emotion. Experiment 3 examined whether temporal parameters influence the perception of emotional expression. The results indicated that emotional intensity and realness judgments were only marginally affected by the speed variations. Experiment 4 compared the amount of information necessary for correctly identifying facial expressions presented either statically or dynamically. The results showed a none significant advantage of dynamic sequences over static ones only for short sequences. This suggests that after a certain amount of information is presented no further benefit is gained. The advantage of dynamic expressions may be due to the contextual information preceding the required judgment Experiment 5 explored this hypothesis via a priming paradigm. The static primes affected the ambiguous test which was judged opposite to them. Dynamic primes showed no effect. Taken together these findings caution the use of static material in postulating underlying perceptual mechanisms associated with emotional facial. This series of experiments supports the ecological perspective by showing that dynamic facial expressions contain relevant veridical contextual information which, as opposed to static facial expressions, is relatively immune to non veridical contextual influence