This study explores the potential of automated linguistic and vocalic analysis tools to identify potentially fraudulent utterances in corporate earnings conference calls. the researchers examine the language and vocal characteristics of restatement-relevant utterances that were prepared (presentation) or unprepared (Q&A) responses. The study finds that restatement-related utterances differ significantly on many vocal and linguistic dimensions, suggesting the value of such analysis tools in detecting potential fraud.
This research highlights the importance of language and vocal analysis in identifying fraudulent utterances during corporate earnings conference calls. The study demonstrates that automated analysis tools can effectively differentiate between restatement-relevant and non-relevant utterances based on distinct linguistic and vocal features. The findings suggest the potential for these tools to improve fraud detection in corporate settings, particularly when analyzing unscripted responses.
Read the research: Which Spoken Language Markers Identify Deception in High-Stakes Settings? Evidence From Earnings Conference Calls
