Age Differences In Emotion Recognition: Task Demands Or Perceptual Dedifferentiation?

ABSTRACT

Background/Study Context: Older adults (OA) have consistently shown lower accuracy compared with younger adults (YA) when labeling facial expressions of emotion in multiple choice tasks. However, OA do not show lower accuracy when judging psychological attributes from faces in rating tasks. The authors investigated whether the cognitive demands of multiple choice tasks yields an underestimation of OA emotion recognition ability and whether lower scores by OA in emotion recognition tasks are an instance of age-related dedifferentiated face perception.

Methods: Younger and older adults judged the emotions of faces depicting various expressions using (a) a multiple choice task and (b) a rating task with separate scales for each expression. We computed both accuracy scores and an emotion differentiation index, adapted from previous work on neural activation to different stimulus categories.

Results: Lower OA performance in emotion recognition do not reflect the cognitive demands of a multiple choice task, since age differences also were shown when assessing emotion expressions on independent rating scales. An index calculating differentiation of the emotion expressions supported the hypothesis that age differences in emotion recognition accuracy may reflect age-related perceptual dedifferentiation.

Discussion: Results are consistent with other evidence for perceptual dedifferentiation in OA, including lower OA performance in face recognition tasks, rating faces more similarly to one another on trait dimensions, and less specificity in neural activation to faces.

ACKNOWLEDGMENTS

We thank Jasmine Boshyan for her help organizing the data collection and Melany Vidret for her help collecting the data.

FUNDING

This research was supported by National Institutes of Health grant AG38375 to the second author.

Notes

¹ For both the multiple choice and rating accuracy analysis, we conducted ANOVAs to examine whether our rater age effects were moderated by face age. No effects were statistically significant (Fs < 1.19, ps > .28).

² Although age differences on the rating task and the differentiation measure argue against a cognitive capacity explanation for age differences in emotion recognition accuracy, we attempted to test this directly by examining whether any of the control measures of perceptual and cognitive function that showed age differences mediated the age differences in emotion recognition. As noted above, OA performed worse than YA on tests of visual acuity, contrast sensitivity, face recognition, processing speed, and recognition of mental states in the eyes. However, none of these measures qualified as mediators of age differences in emotion recognition, as none of the correlations with accuracy were statistically significant when controlling participant age, ps > .25.

Additional information

Funding

This research was supported by National Institutes of Health grant AG38375 to the second author.