Baby Faces: Development and psychometric study of a stimuli set based on babies' emotions.

Background: Sets of stimuli from babies' facial emotions provide a good instrument to detect the recognition of facial emotion (RFE) in clinical and non clinical groups. However, specificities from the stimuli have not been widely explored and validated by previous studies.

New Method: We presented a new set of facial stimuli from infants aged 6-12 months, both sexes, different races, representing five basic emotions. We also present the psychometric properties of validity/reliability for each stimulus and assess whether the sociodemographic characteristics of the stimuli and the subjects affect the RFE.

Results: The stimuli were obtained by a standardized protocol of activities to elicit emotions and 72 stimuli were developed. A total of 119 subjects from the community were selected for the psychometric analysis of the stimuli. The set produced indicators of validity (mean 62.5%) and reliability. Stimuli were evaluated using the Rash model and 15 stimuli had indicators of unpredictability and unmodeled residuals. The difficulty index of each stimulus was calculated, evidencing that the set was normally distributed.

Comparison With Existing Method: Previously published methods are limited in terms of racial diversity, standardisation of the elicitation of emotions, procedure of stimuli extraction, and psychometric evidence.

Conclusions: The findings reinforced the Differential Emotion Theory regarding the expression of basic emotions in infants and evidenced the effect of education level on emotion recognition to the detriment of other sociocultural characteristics (sex and race). This set is freely accessible by email request.