Context shapes evaluation of emotional valence, not emotional categorization: A fresh look at the Kuleshov effect

In filmmaking, a neutral face paired with an emotional context can convey a congruent emotion-a phenomenon known as the Kuleshov effect. However, past research has yielded mixed findings on the existence of the Kuleshov effect, possibly due to methodological variability; we therefore implemented a boundary-test paradigm that combined normed static context images with dynamic facial stimuli separated by an explicit context-rating step to examine how visual context shapes the evaluation and categorization of faces under such conditions. We hypothesized that positive context would elevate facial valence ratings (and vice versa), while evoking different-than-neutral emotions in a neutral face when categorized. Thirty-two participants first rated the valence of a context image on a 7-point Likert scale. They then viewed a video of a neutral face, evaluated its valence on the same scale, and explicitly categorized the expression by selecting one of five predefined emotion labels (neutral, angry, happy, sad, disgusted). Analyses using linear mixed-effects models confirmed that positive contexts led to higher facial valence ratings, while negative contexts led to lower ones; by contrast, multinomial regression revealed no effect of context on emotional categorization. Our findings suggest that while visual context can bias evaluative judgments of facial expressions, it does not necessarily alter their emotional categorization under conditions where cinematic continuity is absent, highlighting the boundaries of the Kuleshov effect in line with previous studies.