Hostile and Aggressive Semantic Themes in Children’s Speech: Discovering Linguistic Indicators for Aggression Restricted; Files Only

Abstract

Specific and targeted indicators and predictors of psychopathology help identify individuals who are at risk for developing mental health problems. The present study investigated whether aggressive children have special linguistic indicators in their speech, reflecting their deficits in social perception. Data were collected on children’s answers to a specific set of open-ended questions after watching video clips of one child ruining another child’s play materials with varying intent. We transcribed children’s recorded speech into text, and used a natural language processing tool to extract instances of noun, verb and adjective use. Then, we manually curated and used ChatGPT to generate aggression-related words and prosocial words. We predicted and found that children who are high in aggression present more semantic themes related to hostility, aggression, and threat - but not prosocial themes - than children who are low in aggression. The results for aggression-related themes held for manually curated but not ChatGPT generated themes. The results should encourage semantic analyses for understanding the causes and mechanisms underlying aggression and for the diagnosis, prevention and treatment of externalizing psychopathology. The results also raise questions about limitations of applying ChatGPT in psychology research. 

Table of Contents

Chapter 1: Introduction...........................................................................1

Chapter 2: Methods.................................................................................5

Chapter 3: Results.................................................................................11

Chapter 4: Discussion............................................................................15

Chapter 5: Appendix.............................................................................20

Chapter 6: Reference.............................................................................30 

About this Honors Thesis

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Emory College
Department	Psychology
Degree	B.S.
Submission	Honors Thesis
Language	English
Research Field	Psychology, Clinical Psychology, General Language, Linguistics
Parola chiave	Linguistic Biomarker Semantics Children's Aggression ChatGPT Externalizing Psychopathology
Committee Chair / Thesis Advisor	Waldman, Irwin, Emory University
Committee Members	Wolff, Phillip, Emory University Choi, Jinho, Emory University

Ultima modifica

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	File download under embargo until 19 May 2025	2023-04-10 11:12:05 -0400	File download under embargo until 19 May 2025

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions