Not All Nudges Land: Behavioral Controllability and Elaboration Quality in AI-Supported Journaling

2026-08-12

AI Summary

This study investigates the effectiveness of AI-delivered behavioral nudges in digital diaries across 26 everyday behaviors, with a focus on whether actions under individual volitional control are more responsive to AI prompts. Introducing the degree of interpersonal dependency as a novel boundary condition, the research integrates passively sensed behavioral data with large language model (LLM)-generated intent annotations. Using a three-day pre-post intervention design and multidimensional textual feature analysis, the findings reveal that behaviors executable independently show significantly higher improvement rates (50–63%) compared to those requiring others’ involvement (15–22%). Although overall textual expression quality exhibits limited predictive power, it demonstrates modest utility specifically in SMS usage patterns and personalized intent articulation.

Abstract

AI journaling tools can tailor prompts to a person's own sensed behavior, but it is unclear which behaviors respond to them. We analyzed 369 journal entries from an eight-week passive sensing study. An LLM labeled each entry as expressing an intention to change a behavior or not, and we measured follow-through against 26 sensor features with a 3-day before/after comparison. Responsiveness depended most on whether a behavior involves other people. Behaviors that depend on others improved in only 15 to 22% of cases, while behaviors a person can act on alone improved more often, up to 50 to 63%, though unevenly. How users wrote mattered less. No single text feature separated improved from unimproved entries; writing carried signal only within specific behaviors, most clearly for text messaging and for longer, more personal intention entries. The sample is small, so we treat these as exploratory patterns that point to where AI journaling nudges are most likely to work.

Problem

• behavioral controllability
• AI-supported journaling
• nudges
• behavior change
• passive sensing

Innovation

• behavioral controllability
• AI-supported journaling
• passive sensing
• large language models
• nudge effectiveness