Participant Information & Consent

PARTICIPANT INFORMATION SHEET

Study Title: AI Work Style Assessment: Understanding AI Delegation and Oversight Patterns Across Occupations

Matthew Beane

mattbeane@ucsb.edu

Principal Investigator

1. Invitation to Participate

You are being invited to participate in a research study conducted by Matthew Beane at UC Santa Barbara. Your participation in this study is entirely voluntary. Please read this information sheet carefully and ask any questions before deciding whether to participate.

2. Purpose

The purpose of this study is to understand how workers across different occupations use AI tools in their daily work — including what tasks they delegate to AI, how much oversight they maintain, and how these patterns vary across professions. This research aims to inform how we think about the evolving relationship between human workers and AI systems.

3. What Participation Involves

• Complete a brief survey (approximately 5 minutes) about your demographics, occupation, and experience using AI tools.
• Export and upload your AI conversation history from one or more platforms you use (such as ChatGPT, Claude, or Gemini). We will provide step-by-step instructions for how to do this. Estimated time: 5–10 minutes.
• Review extracted metadata before it is shared with the research team. Your uploaded conversational data will be processed entirely on your own device. The processing extracts only structured metadata — such as the general category of task performed, approximate effort involved, and indicators of task completion — and does not retain the content of your conversations. You will have the opportunity to review, edit, and remove any metadata entries before consenting to share them. Estimated time: 5 minutes.
• Total estimated time commitment: 15–25 minutes.

4. Session Collector (Optional)

If you use Claude Code, Gemini CLI, or Codex for your work, you may optionally choose to run Session Collector. When run, it extracts the coding session records already stored on your device into a sanitized JSON file — automatically redacting API keys, emails, private IP addresses, home directory paths, and other secrets in the process. Nothing is uploaded automatically: you review that file yourself and can then choose to upload it so it's included in the study's analysis. This is entirely optional, but including it will make your AI Work Style Profile more accurate.

5. Risks

• Discomfort reviewing AI interactions: You may encounter conversation topics in your AI history that you find sensitive and/or uncomfortable upon review. You are under no obligation to include any conversation data that you do not wish to share, and you may remove any metadata entries during the review step.
• Residual privacy risk: Although your raw conversation data is processed locally and never transmitted to the research team, the structured metadata (e.g., task categories, effort indicators) could theoretically be combined with other information to re-identify you. We minimize this risk through data abstraction, de-identification procedures, and secure storage practices described below.

6. Benefits

• While there are no direct benefits to you, you may appreciate receiving a personalized AI Work Style Profile based on your interaction patterns, showing how your AI collaboration style compares to others in your occupation and more broadly. This profile is yours to keep and share if you choose.
• Your participation contributes to scientific understanding of how AI is being used across occupations, which may inform workforce development policies and AI design.

7. Privacy and Confidentiality

• Client-side processing: Your raw AI conversation data is processed entirely on your own device. The full text of your conversations is never transmitted to or stored by the research team.
• Participant review: You review all extracted metadata before it is shared and may edit or remove any entries.
• De-identification: The metadata collected does not include your name, email address, or other direct identifiers, and is linked to your responses using a randomly generated SHA-256 study ID. If you choose to provide an email address — for example, to opt into a follow-up interview or a reminder to finish later — that email is linked to your study ID for that purpose only, and is not included in the study metadata used for analysis.
• Secure storage: All research data is stored on encrypted cloud storage platform.

8. Compensation

You will receive your personalized AI Work Style Profile as a benefit of participation.

9. Voluntary Participation and Withdrawal

Your participation in this study is completely voluntary. You may choose not to participate or may withdraw at any time without penalty. To withdraw during the study, simply close your browser window. To request the destruction of data you have already submitted, contact the research team using the information below.

10. Questions and Contact Information

If you have questions about this study, please contact Matt Beane, Ph.D. at mattbeane@ucsb.edu. If you have questions about your rights as a research participant, you may contact the UCSB Human Subjects Committee at (805) 893-3807 or hsc@research.ucsb.edu.

11. Consent

By checking the box below and clicking "Begin study," you acknowledge that you have read and understood the information contained in this document, that your participation is voluntary, and that you consent to participate in this study.