We will use this space each week in the news blast for Justin to share new ideas he thinks may be of interest for our behavioral & experimental community. If you ever have ideas for topics, please share your ideas with
This Week's BRITE Idea:
For this week's BRITE Ideas we want to highlight a new working paper by Gathergood, Mahoney, Stewart and Weber titled "
How Do Individuals Repay Their Debt? The Balance-Matching Heuristic
". The authors analyze a large data set on credit cards and see how people allocate their monthly payments between cards. The financially efficient thing to do is to allocate payments above the minimum-required payment to the card with the highest interest rate. But people do not do that and are not responsive to the interest rates on their cards.
The main contribution of the paper is that the authors explore a few different heuristics people might be using in deciding how to allocate their payments. They find that a "balance-matching heuristic" where you allocate your share of total payments to each card in proportion to the total balance on that card best fits the data. The reason we wanted to highlight this paper is that it is a very nice example of trying to carefully compare the goodness of fit of different possible decision heuristics. Understanding the heuristics people are using can be a difficult task and is not as straightforward as estimating a parameter from an assumed model -- in these cases you are trying to compare
The really interesting thing in the paper, at least in Justin's opinion, is the way they use some basic machine learning techniques to help explore the models that best predict choice. Justin's opinion is that this use of machine learning is likely to become very popular in behavioral/experimental research, especially as people try to better understand descriptive models of how people make decisions.