Watching the news today, it is easy to think that most people must be dishonest. We see story after story about the latest instance of adultery, or tax evasion, or academic fraud. As I write this, the National League MVP for 2011, Ryan Braun, has finally admitted to cheating in baseball by taking performance-enhancing drugs.

A quick read of the psychology literature may not seem to paint any better of a picture about our characters. For instance, in my last post I mentioned a study by the psychologists Edward Diener at the University of Illinois and Mark Wallbom at the University of Washington where participants were taking a test with the explicit instruction to not go more than five minutes. The experimenter then set a timer bell for five minutes, left the room, and watched what happened through a two-way mirror. 71% of participants kept going after the bell sounded – hardly a display of honest behavior on their part (1976: 110).

But the more I read studies about cheating, the more I am convinced that – for most of us at least – our characters are not so bad after all, and in particular they are not best described as being dishonest. For instance, I neglected to mention that in the same study, there was also a separate group of participants, where each one would take the same test in the same room, except that this time the person was seated directly in front of a two-way mirror and, “thus saw themselves whenever they glanced up” (109). The result? Only 7% of participants cheated (110). This is a startling difference – 71% versus 7%, where the only difference is the presence of a mirror in front of the participants.

Or consider the work of Lisa Shu at Northwestern University and her colleagues (2011), which I also mentioned in my last post. As a reminder, each participant would receive $10 and a worksheet to complete with 20 problems. Each correct answer would earn them $0.50. If you are a participant in the control group, you have no opportunity to cheat, and so get paid for exactly how well you did. If you are in the “shredder” group, you can be sure that you can write down whatever number of correct answers you want to and get paid accordingly, without any possibility of detection. Here were the results:

No Opportunity to Cheat                    7.97 problems answered correctly
                                                            (group average)

Opportunity to Cheat                          13.22 problems answered correctly
                                                            (group average)

Clearly some cheating was going on in the second group. But Shu also did another version of this study, where she added a third group of participants who first signed an honor code before being in the shredder condition. Look at the difference this made (2011: 341):

No Opportunity to Cheat        7.79 problems answered correctly
                                                (group average)

Opportunity to Cheat

            No Honor Code          13.09   problems answered correctly 
                                                (group average)

            Signed Honor Code    7.91 problems answered correctly
                                                (group average)

In other words, cheating seemed to have all but disappeared, even when participants could have made a lot more money by cheating and knew they would get away with it!

Here is a final thing to note. In both of these Shu studies, the group average in the regular cheating situation was around 13 answers. But that was out of 20 question. And this should puzzle us. For why is it that people who knew they could get away with cheating, and had a real financial incentive to cheat, didn’t just say that they got all the questions correct and walk away with 10 dollars?

So here are three respects, then, in which we can see people behaving in ways that are not what I would expect from a dishonest person – by and large, they did not cheat when a mirror is present, they did not cheat when they signed an honor code, and even in the worst case when they did cheat, they did not do so nearly as much as they could have. 

Now don’t get me wrong. I am not saying that therefore most people are honest. As I said in my last post, I think that the experimental work gives us good reason to believe most people are not honest. And now I am saying that most people are not dishonest either.

So this leaves me with two questions:

1)     Does it even make sense to say that most of us are not honest and also not dishonest?

2)     Why do you think the presence of a mirror would make such a difference? Might the explanation have something to do with why signing an honor code also mattered, and with why participants seemed to rarely cheat as much as they could have?

I will consider these questions in the next few posts, but in the meantime please feel free to share your thoughts in the comments.

About the Authors

Christian Miller, Ph.D.

Christian Miller, Ph.D., is a professor of philosophy at Wake Forest University.

Blaine Fowers Ph.D.

Blaine Fowers, Ph.D., is a Professor in the Department of Educational and Psychological Studies at the University of Miami.

You are reading

Questions of Character

Work-Life Balance: Luxury or Ethical Necessity?

The ethical price of overworked America

What Makes a Person Kind?

Exploring the debate between inborn personality traits and situational effects

Why We Love Virtue

(Even If We Don’t Know We Do)