UNVEILING ATTACHMENT STYLES USING AI-GENERATED IMAGES AND NATURAL LANGUAGE PROCESSING (NLP)

Abstract

The current study was intended to unveiling attachment styles using AI-generated images and Natural Language Processing (NLP). Population of the study were married couples. Sample of the study was comprised of N= 30 participants (n=15 couples including, 15 males and 15 females) with age range of in between 25 to 45 years. Couples with marital duration of 3 to 15 years included in the study. Interpretivist epistemology was used in this study. Qualitative research design was used for exploring attachment styles. The study was conducted to develop and innovative method for assessing attachment styles using method for assessing attachment styles using AI-generated images and narrative analysis for analyzing themes of stories and to investigate the relationship between AI-generated images, participant’s created stories and their attachment styles explored through analysis of their stories.Objectives of the study was to evaluate the accuracy of Natural Language Processing (NLP) in identification of attachment styles from participant’s generated stories. AI-generated images was administered to recruit data in the form of stories. Further, analysis for themes extraction was conducted by using Natural Language Processing (NLP). Results from two AI tools: ChatGPT and Claude explored main theme, sentiments, attachment styles, inferred relationships and possible relationship of caregivers in childhood with the participant’s from each story of participants.