Summary of "Week-7.2: Nudges"

Summary of "Week-7.2: Nudges" (Privacy and Security in Online Social Media)

This lecture segment focuses on the concept of Nudges—technological interventions designed to help users make better privacy decisions on social media and online platforms. It highlights the challenges users face in managing their privacy, especially the impracticality of reading lengthy Privacy Policies, and explores experimental tools that aim to improve user awareness and behavior regarding online disclosures.

Main Ideas and Concepts

Privacy Policies Are Rarely Read:
- Most users do not read Privacy Policies due to their length and complexity.
- A US study estimated it would take 244 hours per year per person to read all Privacy Policies of visited websites.
- The national opportunity cost of reading Privacy Policies in the US was estimated at $781 billion annually, illustrating the impracticality of expecting users to read them.
Goal of Nudges:
- To help users make informed decisions about their online disclosures.
- To reduce regrettable online disclosures by providing timely, contextual information.
- To present information in a usable, accessible way rather than overwhelming users with dense policies.
Example Technology from MIT:
- A system that shows users the profile pictures of recipients before sending emails to mailing lists.
- Helps users understand who will receive their message, preventing accidental sharing with unintended audiences.
- Includes a feature that highlights people the user interacts with more frequently, aiding judgment.
Facebook Posting Nudges (Chrome Extension Study):
- A Chrome Extension was developed to nudge Facebook users when posting content.
- Three types of Nudges were tested:
  1. Picture Nudge: Shows profile pictures of people who can see the post (friends, friends of friends, or public).
  2. Timer Nudge: Gives users a 10-second window to cancel or reconsider their post before it goes live.
  3. Sentiment Nudge: Analyzes the sentiment of the post (e.g., negative/angry) and warns users about potential negative perceptions.
Study Methodology:
- Participants: 21 completed the field study; 13 participated in follow-up interviews.
- Data collection via Chrome browser extension and exit surveys.
- Institutional Review Board (IRB) approval ensured ethical treatment of participants.
- Metrics analyzed included:
  - Changes in privacy settings.
  - Number of canceled or edited posts.
  - Post frequency.
  - Sensitivity of topics posted.
Key Findings:
- Picture Nudge:
  - Some users changed their audience settings to more restrictive groups.
  - Some canceled posts after realizing more people would see them than expected.
- Timer Nudge:
  - Mixed reactions: some found it annoying, others found it helpful.
  - Encouraged users to think before posting.
- Sentiment Nudge:
  - Contextual understanding of sentiment was limited and sometimes inaccurate.
  - Many users canceled posts flagged as negative.
  - Overall post frequency decreased (from 13 to 7 posts on average).
- Nudges were effective in helping users reconsider posts that might offend or be shared with unintended audiences.
- More research is needed to refine Nudges and adapt them to different contexts (e.g., frequent posting scenarios).
Conclusion and Future Directions:
- Nudges show promise in improving privacy decision-making on social media.
- The design and context of Nudges need further development to balance usability and effectiveness.
- The instructor encourages students to explore this area for potential research projects.
- The next lecture will focus on phishing in social media contexts.

Detailed Bullet Points: Methodology of the Facebook Nudge Study

Developed a Chrome browser extension to implement Nudges on Facebook posts.
Three nudge types tested:
- Picture Nudge (shows audience profile pictures).
- Timer nudge (10-second delay before posting).
- Sentiment Nudge (sentiment analysis warning).
Participants recruited via flyers, emails, Craigslist, etc.
21 participants completed the field study; 13 participated in exit interviews.
Data collected included:
- Number of posts canceled or edited.
- Changes in privacy settings.
- Post frequency before and after Nudges.
- Sensitivity of post topics.
Ethical approval (IRB) obtained to protect participant rights and privacy.
Both quantitative data and qualitative feedback (interviews) were analyzed.

Speakers/Sources Featured

Primary Speaker: Course instructor (unnamed) delivering the lecture on Privacy and Security in Online Social Media.
Researchers Referenced:
- MIT researchers who developed the email recipient visualization tool.
- Researchers who conducted the Facebook nudge Chrome Extension study.
Mentioned Studies:
- US study on time and economic cost of reading Privacy Policies (PhD thesis work).
Examples: