Txt2Stop: A Behavior Change Technique Analysis of Smoking Cessation Messages

899 text messages from the 'Txt2Stop' smoking cessation program were coded using a taxonomy of 34 behavior change techniques (BCTs). The content focuses heavily on enhancing self-regulatory capacity (870 messages) and maintaining engagement (552 messages), with less emphasis on motivation (218 messages) and adjuvant behaviors (39 messages). This dataset provides a standardized characterization of intervention content for comparative effectiveness research in digital health.

Use Cases

• Comparing the BCT composition of different digital smoking cessation interventions based on a standardized taxonomy.
• Analyzing the relationship between specific BCTs (e.g., self-regulation techniques) and intervention outcomes in clinical trials.
• Developing new text-message interventions by modeling the BCT profile of an evidence-based program like Txt2Stop.
• Training machine learning models to automatically code behavior change techniques in health communication text.

Strengths

• Provides detailed counts of 899 messages categorized by specific behavior change technique.
• Uses a defined taxonomy of 34 BCTs adapted from face-to-face counseling, offering a structured coding framework.
• Cross-platform presence on PapersWithCode signals its recognized use in computational research.

Limitations

• Key metadata is missing: no column names, row count, or last update date are provided in any source.
• The dataset has a 'closed' license, restricting access and reuse.
• Inter-rater reliability for the coding system was reported as moderate (61%), indicating potential subjectivity in annotations.

Provenance

Source

University College London (author: Susan Michie)

Collection Method

Content analysis of all messages in the Txt2Stop SMS program using an adapted BCT taxonomy.