This study is part of a research program that aims to explore the influence of emotional state on human bodily expression. The current project focuses on motion parameters related to human walking and on upper body parameters. Two experiments were conducted in order to identify motion parameters that might vary from one emotional state to another.

In the first experiment 40 undergraduate students from the Department of Management at Ben-Gurion University of the Negev, watched four different three-minute movies segments, which induced each one of four emotional states:​ happiness, sadness, relaxation and fear, which are divided into a two-dimension system which divide emotional states by their level of arousal (high arousal vs low arousal) and valance (positive vs negative). Participant’s posture and motion parameters were captured during standing and walking positions, and recorded by QTM (qualysis track manager) system, comprised of 8 cameras, and 51 reflectors located on the various points of the participants’ body. During the experiment and between each movie segment, cultural orientation, emotional regulation tendencies, personality traits, and demographics data was collected via different questionnaires, in order to uncover possible moderating roles of various individual differences between emotional state and motion. Motion parameters such as walking speed, stride length and different body angles, were calculated via a custom made algorithm on Matlab software and based on 3-diementional coordinates delivered by QTM system, and was divided into 3 groups of parameters:​ tempo of body motion, body posture and stability of motion.

Results of the first experiment indicate for greater body motion under happiness condition (highest walking speed and stride), and more upright shoulders posture for this condition (higher shoulder mean angles) while relaxation condition is featured with the lowest values for these parameters. Fear is characterized by higher standard deviation which imply less stability compared to relaxation condition. Moreover, neck and back angle results show that for sadness and relaxation conditions (low arousal), lower mean angle was spotted which indicates more upright back and neck posture for these parameters compared with happiness and fear conditions.

In second experiment 46 participants performed the same procedure as in the first experiment. Participants were classified to one of two cultural orientation groups based on their responses to a cultural orientation scale which was administered in a pretest study (individualists or collectivists). Participants’ posture and motion parameters while standing and walking were captured using Vicon System, comprised of 16 cameras and 17 reflectors. Data creation (Matlab custom-made code) and analysis were performed similarly to the first experiment.

Results within the second experiment indicate greater rhythm of body motion under happiness condition compared to sadness, which featured with lower rhythm of body motion (walking speed and stride). Back angle results imply for better upright posture under happiness and fear conditions compared to more bent posture under sadness condition. Cultural orientation tends to have a moderating role between the emotional state and body motion parameters, as results show different patterns of behavior among individualists and collectivists. This influence appeared significant for some of the motion parameters (e.g. walking speed and stride) while for shoulders angle, influence proved insignificant.

This research may have implication in a variety of fields. One example is the field of emotion detection. This research shows that unintentional expression of body motion can be captured and analyzed in order to detect different emotional states. Moreover, the role of additional moderating factors such as gender and the cultural orientation of a person have been shown to have possible effects on body motion and therefore should be taken into equation when developing classifiers to distinguish between the emotional states.