Search
My independent study: the experiment
In Part 1, I talked about the question that drove my independent study. In short, my question was this: why should people experience loss aversion for an object they give up, but not for money that they give up?
Given this supposed difference between money and objects, studies of the endowment effect potentially confound an increase in value upon ownership with a simple difference in the form of what is owned and what is not owned. To get around that confusion, I needed to replace money with another object. And to get precise prices, that object needed to be easily conceptually divisible. Previous studies had used gourmet chocolate bars in place of coffee mugs, so I decided to use chocolate bars to replace money in my experiment. Like money, chocolate bars can easily be broken into smaller, defined pieces, but unlike money, they have value unto themselves, so they should be subject to the same terms of loss aversion as a coffee mug. HypothesisIf, even when trading two objects, I found that the selling price was higher than the buying price for the mug, then I could reject the loss aversion hypothesis since people should have experienced the same loss aversion for the chocolate as for the mug, negating the effect. But if the gap disappeared or got smaller, then I could say that the loss aversion explanation is plausible, with the added clause that money is somehow exempt from loss aversion in these situations.ProcedureBut before I could get ahead of myself and dive straight into my manipulation, I had to make sure that I could replicate the standard endowment effect. If I would not find the effect when people payed for the mug with money, how could I say anything about my contribution?You can read the full procedure in my paper (linked below), but for now all you need to know is the following.
- 100 participants, 50 each in the control (money) experiment and the test (chocolate) experiment
- I "gave" half the participants in each experiment a mug from the campus bookstore. These were the sellers
- I showed the other half the mug, but did not give it to them. These were the buyers.
- In the money experiment, I asked the sellers a series of questions meant to find the point at which they were indifferent between their mug and some amount of money, or their willingness to accept (WTA). I asked the buyers similar questions to find their willingness to pay (WTP).
- In the chocolate experiment, I also "gave" the buyers some chocolate bars.
- In the chocolate experiment, I asked similar questions to find sellers WTA and buyers WTP in terms of number of pieces of chocolate.
he quotes here mean that the action within is hypothetical. Participants didn't really own the mug or the chocolate, and no actual trades were made as a result of their price information. My paper explains this decision further.
ResultsIn short, yikes! Kind of a mess. I said that I had to replicate the standard endowment effect if I wanted to say anything about the chocolate experiment. And unfortunately, I could not. The selling price was higher than the buying price, but not by a statistically significant amount, so there was no effect of endowment in my replication. Also, I learned that my hypothesis was wrong, not just in terms of what I thought would happen, but in terms of what that would mean. I'll go into both of these in the next post, since the implications are important. But I'll leave you a small cliff off of which to hang: if loss aversion is what is going on, then the gap between the selling and buying price should be bigger when trading two objects, not smaller.
