Reduced willingness to approach genuine smilers in social anxiety explained by potential for social evaluation, not misperception of smile authenticity

ABSTRACT

We investigate perception of, and responses to, facial expression authenticity for the first time in social anxiety, testing genuine and polite smiles. Experiment 1 (N = 141) found perception of smile authenticity was unaffected, but that approach ratings, which are known to be reduced in social anxiety for happy faces, are more strongly reduced for genuine than polite smiles. Moreover, we found an independent contribution of social anxiety to approach ratings, over and above general negative affect (state/trait anxiety, depression), only for genuine smiles, and not for polite ones. We argue this pattern of results can be explained by genuine smilers signalling greater potential for interaction – and thus greater potential for the scrutiny that is feared in social anxiety – than polite smiles. Experiment 2 established that, relative to polite smilers, genuine smilers are indeed perceived as friendlier and likely to want to talk for longer if approached. Critically, the degree to which individual face items were perceived as wanting to interact correlated strongly with the amount that social anxiety reduced willingness to approach in Experiment 1. We conclude it is the potential for social evaluation and scrutiny signalled by happy expressions, rather than their positive valence, that is important in social anxiety.

KEYWORDS:

• Facial expression
• genuine
• Duchenne
• social anxiety
• positive valence

Acknowledgements

We thank Tamara Gradden for help with testing some participants and Dr Rachel Robbins for assistance with preparing data and figures.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Amy Dawel http://orcid.org/0000-0001-6668-3121

Richard O’Kearney http://orcid.org/0000-0002-7839-7920

Elinor McKone http://orcid.org/0000-0003-1655-4297

Notes

1 The sample overlaps with the N = 94 “average observer” sample Dawel et al. (2017, Experiment 3) and N = 140 from Dawel et al. (in press). Those articles do not include the key measures reported here (social anxiety, approach ratings).

2 The main effect of social anxiety group was also significant, F(1,139) = 8.03, MSE = 2.70, p = .005, ${\eta }_p^2$ = .06, but was uninformative in light of the significant interaction and t-test results. There was no significant main effect of smile type, F(1,139) = 0.43, MSE = 0.70, p = .514, ${\eta }_p^2$ < .01.

3 Intensity differences are difficult to avoid because intensity is in fact one of the physical cues that can contribute to perceiving a happy expression as genuine; Dawel et al., 2015).

Additional information

Funding

This work was supported by the Australian Research Council (DP110100850; CE110001021, see ARC Centre of Excellence for Cognition and Its Disorders at www.ccd.edu.au).