Socially anxious individuals discriminate better between angry and neutral faces, particularly when using low spatial frequency information

Background and objectives: Social anxiety is associated with biased processing of threatening faces. Earlier research indicated that socially anxious individuals are biased towards processing low spatial frequency (LSF) information when judging facial expressions. However, it remains unclear whether this bias reflects better performance for LSF-information, worse performance for high spatial frequency (HSF) information that needs to be compensated for, or both.

Methods: To answer this question, we used frequency-filtered neutral and angry face stimuli in a speeded classification task to compare the performance of socially anxious and non-anxious individuals for different spatial frequency bands.

Results: Across all spatial frequency bands, socially anxious individuals were faster in judging facial expressions. Importantly, this performance advantage was larger for LSF-filtered stimuli and most pronounced for those stimuli with the lowest frequency band. Analyzing inverse efficiency scores showed the same pattern, ruling out speed-accuracy trade-off differences between groups.

Limitations: The study uses rather artificial (bandpass-filtered) stimuli and is limited towards contrasting the discrimination of neutral and angry faces. Further, only participants with subclinical anxiety were part of the study, so clinical relevance remains to be shown.

Conclusions: Our results show that social anxiety is not characterized by deficits in judging emotions from HSF-information, but by advantages when processing LSF-information.