Objective: To evaluate the impact of facial lesions on affect display in patients as compared with individuals without lesions.
Study design: Prospective randomized controlled experiment. Setting Academic tertiary referral center.
Subjects and methods: Forty-five normal observers viewed photographs of normal faces in repose and faces with lesions in repose. Observers classified the affect display of the patients and the normals using a survey containing choices of primary emotions and personal attributes. Latent class analysis was used to find the unmeasured or "latent" classes. Latent class regression was used to determine the impact of lesion size and location on class membership. Finally, lesion classes and normal classes were checked for differences.
Results: Latent class analysis identified 4 independent latent classes for faces with lesions. The normal data were best fit by 3 latent classes. As a group, faces with lesions were most likely to be classified as either negative-sad or negative-angry, accounting for a combined probability of 44%. This was in contrast to the normal group, where 62% were classified as positive and only 13.9% as negative. Size and location significantly affected negative class membership. There were significant differences in the lesion classes as compared with the normal classes.
Conclusion: Faces with lesions were reported to display different affect in repose than normal faces in repose. The classification of the lesion faces was dependent on lesion size and location. These findings provide the first objective evidence of the affect display penalty caused by facial lesions.
Keywords: facial attention; facial deformity; facial perception.