Do emotions result in their predicted facial expressions? A meta-analysis of studies on the co-occurrence of expression and emotion.
Level 1 - systematic review of randomized trials
Meta-analysis (graded by design analogy for non-clinical research)
PubMed 34780241 · doi:10.1037/emo0001015
What was done
Random-effects meta-analyses evaluated the co-occurrence of six basic emotions (happiness, sadness, anger, disgust, fear, surprise) with their predicted facial configurations across three analyses: (1) studies measuring full facial expressions using the Facial Action Coding System (FACS); (2) studies measuring full or partial expressions using FACS or alternative systems; and (3) studies assessing the Pearson correlation between the self-reported intensity of an emotion and the intensity of the predicted facial expression.
What was found
The average co-occurrence effect size was 0.13 for full expressions measured via FACS and 0.23 when partial expressions and other measurement systems were included. The average Pearson correlation between reported emotion intensity and expression intensity was r = .30. Co-occurrence and correlation were highest for disgust and lowest for surprise. Significant between-sample heterogeneity was observed.
Why it matters
This review challenges the foundational assumption of classic basic emotion theory, demonstrating that distinct emotional states do not reliably produce stereotyped facial expressions.
Limits
The abstract does not report the total number of included studies, sample sizes, or participant demographics. Substantial between-sample heterogeneity was present, and findings depend on varying expression coding standards.
Cited by
- supports Motor patterns of emotional expression and the verbal language used to describe emotional experiences correlate weakly at approximately r = 0.2.