The representational momentum effect and the reference dependence effect on the evaluation of happy expressions

The majority of current research on facial expression perception uses static face images, and there is comparatively little study on dynamic expression. People's expression in daily life is frequently a dynamic process, and its temporal characteristic will also provide more details. Yoshikawa and Sato (2008) noted that when individuals observe dynamic expressions, they would experience the representational momentum effect; however, it is still unclear how changing direction, one of the fundamental signals of dynamic expressions, will affect the effect. In addition, researchers also need to put specific experimental designs and materials into the study of specific expressions, so how different experimental designs affect people's emotional perception results of certain expressions.Three experiments had 96 participants in all, and 94 valid data were finally collected. The experimental material was from the CASP, and then the happy and neutral expressions of the same identity were made into dynamic expression sequences. Experiment 1 explored the effect of changing direction and the average summary representation on emotion perception outcomes. Experiments 2a and 2b systematically explored the representational momentum effect of dynamic happy expressions by combining static expressions, the difference was that the average summary representation of dynamic expressions in the two experiments was different. On a seven-point scale, participants were asked to evaluate the valence, arousal, and dominance of final dynamic emotions or static expressions. Finally, the data were analyzed by ANOVA, independent sample t-test and one sample t-test.Experiment 1 found that compared to dynamic happy emotions with intensity changed from strong to weak, those with intensity changed from weak to strong were shown to have greater valence scores and lower dominance scores. The dynamic happy expressions with higher the average summary representation was considered to have higher arousal scores. Experiments 2a and 2b found the representation momentum effect when people perceived the valence and dominance of the dynamic happy expressions. Depending on whether the intensity of the expression varied from weak to strong or vice versa, the valence of the dynamic expressions was scored higher or lower than that of the static expressions with the identical expression in the previous frame. The dynamic happy expressions that changed from strong to weak scored higher in the dominance dimension. Dynamic expressions elicited more arousal than static expressions. The same expression would be perceived as having varying valence, arousal, and dominance outcomes in different experiments, according to a comprehensive comparison of the three experiments.In conclusion, the results found that people's perception of the valence and dominance of dynamic happy expressions had a representation momentum effect, and the perception results would be further extended to the direction of facial expression changes. The arousal of dynamic expressions was always higher than that of static expressions. The same expression will be perceived as different results in different experiments, that is, the reference dependence effect was appeared in expression perception. These processing characteristics remind researchers to pay attention to the difference between dynamic and static expressions, and the influence between materials when using facial expression materials in future research.