Analysing the Context of Nonverbal Behavioural Cues in Multiple Social Settings

Abstract

Socio-cognitive disorders such as depression or social incompetency in communicating and interacting with others are common among all genders. The traditional way of diagnosing these disorders involves the examination of individuals through verbal cues extracted from speech and communication patterns identified by clinicians in clinical settings. However, this is time consuming and costly, and chances of error of clinical judgement is quite high due to the artificial speaking style and wordings used by individuals afflicted by cognitive and social disorders. Therefore, there is need to develop methodologies which can be used to examine the social and mental health conditions of people without their conscious involvement. This can lead to early detection of disorders, followed by implementation of effective and timely therapeutic strategies. This work aims at such a goal by analysing non-verbal cues in different social settings. This study used telephone call data from 120 subjects and interview transcriptions involving 59 individuals (including 30 diagnosed with depression) and analysed the distribution of the context surrounding nonverbal cues, i.e., the type of behaviour that was observed immediately before and immediately after a given nonverbal cue. The results show that there are significant differences in the distribution of silence, lengthening, fillers, laughter, between the dataset depressed individuals and the others, thereby indicating the potential of non-verbal cues as of depression. The analysis of telephone call datasets indicated that silence, fillers and laughter tend to appear in different contexts. The data generated in this study contributes to the existing literature based on corpus analysis methodologies.