Social Spaces on the Internet

Fall 2018

Assignment 2


1.  Catalina L. Toma & Jeffrey T. Hancock. Reading between the Lines:  Linguistic Cues to Deception in Online Dating Profiles. CSCW 2010.
2.  Judith Donath.  Identity and Deception in the Virtual CommunityCommunities in Cyberspace. P. Kollock and M. Smith  (eds). Routledge. 1999.
3. Jeffrey T.  Hancock, Jennifer Thom-Santelli, & Thompson Richie.  Deception and Design:  The Impact of Communication Technology on Lying Behavior. CHI 2004
4. Tanushree Mitra, Graham Wright, & Eric Gilbert. Credibility and the Dynamics of Collective Attention. CSCW 2018.

(Optional) Gordon Pennycook and David Rand. Who Falls for Fake News? The Roles of Analytic Thinking, Motivated Reasoning, Political Ideology, and Bullshit Receptivity. SSRN. 2017.

Write up

1.  Suppose, hypothetically, that you built a lie detector for  (a) email, (b) Skype.  Let's say your phone vibrated whenever a lie was detected.  Discuss the sender and receiver costs of this signal in each of these scenarios.  For each scenario, describe the costs when the systems incorrectly suggest a lie and when they "miss" a lie.

2.  As an alternative to the system described above, instead of explicitly alerting the user that a "lie" occurred, how might you design an interface to encourage "authenticity" in interaction.  Make sketches with descriptions for email and Skype for a specific scenario.  Describe your design choices.  Would this improve the quality of conversation in romantic, business, courtroom, classroom settings?  Compare and contrast to no change in the interface.

3. Do you believe it is beneficial to design real time "lie detectors" into synchronous communication systems (e.g., Skype, Hangouts)?  Do you believe it is beneficial to design lie detectors into socio-technical news/recommender system system?  Why or why not?

4. In face-to-face interaction a sender may sense disbelief in the receiver from their reaction (e.g., facial expression, posture, etc.) In questions 1-3, you considered automated lie detectors. What are the pros and cons of automated lie detectors, versus systems where the receiver signals a lie either directly or indirectly?

5.  Deception is encouraged or expected in some online communities (e.g., online poker).  In other communities, deception can result in higher sender and receiver costs (e.g.,  health forums, Wikipedia, etc.).  Choose two online communities - one where the cost of lying to the receiver is high and one where it is low.  Discuss what, if any punishment policies or norms have evolved in these communities. 

Responses should be submitted via compass in PDF form. 

Late responses will not be accepted.

Academic Integrity
For more information, please review the University Student Conduct Code for Academic Integrity.