Detecting Deception: Uncovering Lies in Asynchronous Video Interviews Using Natural Language Processing

Simple item page
dc.contributor.advisorHell, Jurgen
dc.contributor.authorKrumrey, Jule
dc.date.issued2021-07-26
dc.description.abstractApplicant faking in interviews is a common problem in recruiting. Currently, it remains unclear what influence technology-mediated interviews have on applicants’ faking behaviour. Therefore, this research aims at detecting deceptive interview answers in asynchronous video interviews using structural natural language processing and a closed-vocabulary approach. The between-subjects design compared the answers of three past-behavioural questions between an induced deceptive (n = 48, M = 44%, W = 56%) and an honest group (n = 120, M = 71%, W = 29%), targeting the characteristics word count, first-person pronouns, disfluencies and the number of retries that the candidates used. Independent samples t-test confirmed the hypothesis of more frequent disfluencies in the deceptive group. For retries, the hypothesis could only be confirmed partly. No support was found for the remaining hypotheses. The main limitations include the sample characteristics and the transcription service that were used. A crucial implication is the lack of a high-stake situation which seem to affect the responses. The present paper contributes to the empirical evidence of detecting deception in asynchronous video interviews, although the generality of the findings presented in this study must be ascertained by future research. Keywords: AVI, deception, impression management, interview, asynchronous video interview
dc.identifier.urihttps://theses.ubn.ru.nl/handle/123456789/14533
dc.language.isoen
dc.thesis.facultyFaculteit der Sociale Wetenschappen
dc.thesis.specialisationspecialisations::Faculteit der Sociale Wetenschappen::Psychologie::Master - Arbeid Organisatie en Gezondheid
dc.thesis.studyprogrammestudyprogrammes::Faculteit der Sociale Wetenschappen::Psychologie
dc.thesis.typeMaster
dc.titleDetecting Deception: Uncovering Lies in Asynchronous Video Interviews Using Natural Language Processing
Files
Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
Krumrey-Jule-s1063853-Thesis-2020-21.pdf
Size:
848.28 KB
Format:
Adobe Portable Document Format
Download
Collections
Faculteit der Sociale Wetenschappen