Authors:

Grace Lin

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

As artificial intelligence integrates into daily life, its role in traditionally subjective domains remains underexplored. This study examines how individuals evaluate fortune-telling claims depending on the perceived source (Al versus human) and linguistic complexity (astrology terminology versus mundane language). We hypothesize that Al-attributed messages will be rated higher in trustworthiness and yield greater positive changes in persuasion outcomes than human-attributed messages. We also predict a main effect for linguistic complexity, where technical, fortune-telling-specific language increases trust, alongside an interaction effect: human-attributed messages will be rated dissimilarly across linguistic complexity conditions, while Al-attributed messages will be rated more similarly. In this experimental design, adults who have consulted fortune-telling or astrology at least once in the past year read one of four randomly assigned fortune-telling messages. These messages manipulate source and language while preserving core content. We measure trustworthiness and calculate overall persuasion outcomes using pre-test and post-test change scores for behavioral intention and attitude. Preliminary descriptive data (N = 15) indicates a strong emerging main effect for linguistic complexity: messages using complex astrological terminology yielded substantially higher trustworthiness (M = 6.57) and positive persuasion changes (M = 0.57) compared to mundane language (M = 3.55 and M = -0.21, respectively). While overall trustworthiness was similar across Al (M = 4.43) and human (M = 4.38) sources, an unexpected interaction trend emerged: complex language increased trustworthiness and persuasion more drastically for Al-attributed messages than for human-attributed ones, diverging from our initial prediction. Consequently, these early findings suggest that in fortune-telling contexts, specialized terminology may be critical for lending credibility to artificial intelligence.

Source:

Northwestern University

Topics:

No topics listed

Co-authors:

Grace Lin