Valdez, D., Jozkowski, K.N., Haus, K.R., Crawford, B.L., & ten Thij, M. (2022, March 13 - May 16). Natural Language Processing and Opinion Mining: Assessing abortion belief systems in pilot qualitative interviews about abortion ideology. American Academy of Health Behavior (AAHB) Annual Meeting, Key Largo, FL.
Background. Although Pro-Choice and Pro-Life are synonymous with abortion ideology, empirical assessments of how people communicate personal abortion views is lacking. Advances in big-data science now afford opportunities to study words, or patterns of words, that may denote inflexible and/or rigid ways people communicate beliefs. Purpose. This pilot study applies Natural Language Processing (NLP) methods to compare qualitative interview transcripts about abortion beliefs against a lexicon comprised on n-grams (or a series of words) indicative of cognitive distortions (i.e., markers of inflexible thinking). Methods. We ran interview transcripts against a cognitive distortion lexicon and tallied the number of cognitive distortions spoken per interview divided by length of each interview (for standardization). Results. Our results indicated that people who strongly identified as Pro-Choice or Pro-Life used more cognitive distortions to express their views than people who identified as equally Pro-Choice and Pro-life or did not associate with either label. The three most represented classes of cognitive distortions included (1) dichotomous reasoning (black and white thinking, 30%) (2) mind-reading (assuming people share one's views, 42%) and (3) normative thinking (expressing one's views as morally correct, 15%). Discussion & Conclusion. Our findings suggest that polar abortion views may be grounded in inflexible thinking patterns and limited complexity. By contrast, moderate or uncertain abortion views may be more conflicted and mired in complexity. Implications associated with our findings suggest strong attachment to social issues may hamper the ability to think openly about the issue or consider points contrary to one's own belief system.