For this post I decided to do an analysis of deception in one of my friend’s Facebook profile. Although I did not generally get the impression that people I knew were prone to lying, especially about the superfluous things that one puts on Facebook, I thought it would be interesting to challenge this bit of truth bias in myself. How much can we really trust what our Facebook friends are telling us over the Internet?
I picked one of my non-Cornell friends at random for the study. I wanted to avoid potential bias from a Cornell student who might have heard of this class or the assignment and thus would be less willing to give truthful answers. (Turns out INFO 245 is pretty darn popular around campus; almost everyone I have spoken to is either in the class or knows someone else in the class!) I also did not mention to my subject what exactly I needed help with, saying only that it was for a class assignment, until the person had agreed to help. At this point, I went through his Facebook profile and set up a grid for him to rate each of his profile elements for veracity on a scale from one to five. He entered his answers into the document and mailed it back to me. After receiving it, I cross-referenced his self evaluation with what knowledge I had of him; since he was a friend of mine, that knowledge was fairly extensive in this case.
The hypothesis put forward by Goffman and Baumeister states that people should be tempted to lie frequently, subtly, and in accordance with the wishes of the opposite gender. Based on this information, I should have expected to see many small blurred truths as opposed to bald-faced lies.
What I observed was somewhat consistent with that hypothesis. The subject reported minor to significant lies in over half the elements on his profile. The more significant lies possibly strain the “lie subtly” component of the hypothesis, but I believe that it can be explained by analyzing the types of information that he lied in.
The subject evaluated himself as perfectly accurate for the “basics” section (name, gender, networks, political views, etc; all the items that appear in the top section of Facebook). Due to the nature of Facebook, which tends to unite people in a relatively limited geographic area, it would have been nearly impossible for him to fake these core assessment signals; any lies would be detected instantly by the people sees on a daily basis.
The other sections showed less perfect truths. For the “favorites” (books, music, television shows, etc.), he rated himself four, with five being completely accurate and one being completely inaccurate. In “Interests” and “Quotes,” however, he rated his accuracy at only three and two, respectively. This disparity can probably also be explained in the difference between assessment and conventional signals. In order to claim a particular book or show as a “favorite,” he would have needed to invest substantial time in reading the book or watching the show. This makes the “favorites” list more of an assessment signal. However, for the quotes section, finding quotes on the Internet to fit one’s desired self image is not difficult at all; quotes are therefore a conventional, easily faked signal.
Comments:
Subscribe to:
Post Comments (Atom)
3 comments:
It seems like your findings were fairly similar to most other posts. Assessment signals were almost all correct while conventional signals were commonly lied about. I'm not totally sure what you mean when you reference the Goffman and Baumeister theory. Why would someone’s facebook profile be seen by only the opposite sex or why would a person customize their profile with exactly the opposite sex in mind? I feel like this might be more relevant to online dating sites. I think that facebook allows you to lie even more severely because you’re not necessarily ever going to meet the people who view your facebook profile. I do agree with your mention of favorites as being more of an assessment signal. It could be easy to find out if that is actually someone’s favorite book by asking a few questions. It seems a lot of items on facebook could possibly be construed as assessment or conventional signals. Some are more easy to lie about and they all depend on how close you are to the person you are analyzing.
Hey Chris,
I found your post quite interesting. It was great how you maintained the unbias environment necessary for an experiment to the utmost care.
I however, was confused with what you meant by what Goffman and Baumeister said. I feel that to base your hypothesis on the fact that "people should be tempted to lie frequently, subtly, and in accordance with the wishes of the opposite gender", is not so much of an issue on facebook; I do not feel you would see "many small blurred truth" because of the aforementioned reason. Rather, I feel that it has more to do with just being the one that is "in" with the times. I feel this way because Facebook is more of a social networking forum as opposed to an direct dating forum.
But regardless, I guess that is what one would think, or at least I thought so, but to see that you found people who showed results otherwise....is great!
Great job Chris
Chris-
Great post! Your hypothesis and further observations caused me to think about my own facebook and the degree to which deception is present. I came to the conclusion that facebook might not be the most appropriate medium to analyze for deceptive content due to the face that it really has no legitimacy to begin with. For example, under the personal sections on my facebook profile, I include a lot of things that are personal jokes with friends. I include these on my profile in order to be humorous to the people who I think are going to be looking at my profile. My facebook is private, and I therefore know who is able to look at my profile. In effect, my facebook profile is not really a medium that I use to portray myself to the public. I think a more appropriate medium to evaluate for truthfulness would be a resume-posting website or a dating website, when users are providing posting in order to give off a good impression to complete strangers. An interesting study would be to determine which type of site (facebook, monster, jdate, etc) contains the most deceptive profiles. Good job!
Post a Comment