Do People Cheat During On-Line Surveys?

by John Sides on April 18, 2012 · 4 comments

in Public opinion

Of the 505 people who completed the survey on a computer, only 2 people cheated by looking the answers up on-line.  That’s less than one-half of one percent of the respondents.  This hardly qualifies as an alarming finding.  Or as reason for “word” to spread of cheating on political surveys.  Plenty of people had a hard time answering our fact-based questions, and they knew they were on the Internet, yet very few of them took the time to look up the answers—- in fact, almost none of them.  And, one of the respondents who did look up the VPs name felt so badly about doing so, he came out and told us he cheated.

Lynn Vavreck, writing at Model Politics.  More here.

Why don’t more people cheat?  Part of the explanation, I think, is that people typically don’t want to spend a lot of time answering a poll, even on-line.  (We’ve all taken these polls.  The impulse to click quickly through questions is pretty strong.)  Because of this, they “satisfice” in various ways.  Which means they don’t take the time to look up answers to factual questions on Wikipedia or wherever.

{ 4 comments }

Adam Berinsky April 18, 2012 at 10:02 am

It seems to me that the experience of taking a survey in the privacy of one’s home is different on this score than taking it on a computer that a researcher just directed you to. It’s an open question to be sure and this evidence is quite interesting, but it would like to see more work on this question before I would be willing to close the case. Maybe we could start collecting web browser cache information?

Brendan Nyhan April 19, 2012 at 12:12 pm

I agree with Adam – I’m concerned about this problem, and think behavior in the field (i.e., respondents at home) may differ from what Lynn found. I’ve started putting time limits on political knowledge questions during online experiments (typically 20-30 seconds) to limit respondents’ ability to look up the answers.

soclosetobeingdonewithmyphditskillingme April 19, 2012 at 8:13 pm

i don’t think putting time limits solves anything, and you can always remove outliers later (if you have response latencies).

there’s a funny juxtaposition in survey research–we think respondents don’t care enough to think thoughtfully through our questions, but we also think that they care enough to cheat. look at open-ended responses to survey questions–lots of people don’t know but just give you non-sense remarks.

John April 24, 2012 at 9:41 am

Depending on the type of survey that the person is taking would determine if they would cheat or not. I would say that people do nto cheat on the average consumer survey that does nothing more than gather data with no reinbersment to the individual. The cash surveys that people do to get a few extra bucks a month now thats a different story, I would say that people cheat on those soley for the fact they want to make it past the screening portion so they can get paid.

Comments on this entry are closed.

Previous post:

Next post: