«Building and Rebuilding Trust with Promises and Apologies Eric Schniter and Roman Sheremeta and Daniel Sznycer Online at ...»
Appendix B: Instructions for Coders
You will be playing a MATCHING GAME with two other students, called your partners. Your payment for the matching game depends not only on what you do, but also on what your partners do, and on the rules of the matching game described below. Make sure you understand the rules. Feel free to ask us questions as they arise, by raising your hand. Please do not speak to other students during the experiment. You will receive $7 for participating in this session. You may also receive additional money, depending on the decisions you and the two other students make. Upon completion of the session, this additional amount will be paid to you individually and privately.
Over the last two years, many Chapman students like you came to our lab to participate in economic experiments which we call “trust games.” In these games, each participant was paired with another participant. All participants interacted via computer terminals and the interaction was anonymous. That is, no participant knew the identity of the person with whom they were paired. In each pair, one person had the role of A, and the other had the role of B.
The trust game proceeded in several stages. First, by choosing a dollar amount from $0 to $20, B indicated the proportion of a possible $20 income that he or she promised to transfer back to A, should A choose IN.
Specifically, B completed the following statement: “I (Participant B) promise to transfer back ___ of my income to you (Participant A) if you choose IN”. Both B and A understood these promises to be non-binding. The computer conveyed B’s statement to A. Having received a statement from B, A chose either IN or OUT. If A chose OUT, A received $5 and B received $0. If A chose IN, B received $20 income. In such a case, after receiving $20 income, B chose a dollar amount from $0 to $20 to transfer back to A.
Then we presented exactly the same trust game to the same participants once again. Neither participant A nor participant B knew that there would be an unexpected second trust game. After learning the outcome of the first trust game (from now on we will call it Game 1) and after the unexpected second game (Game 2) was announced, but before it commenced, B could send a one-way message to A – saying anything that did not personally identify them.
You are NOT playing these trust games! Instead, you and your partners will play the matching game.
In the matching game you will be examining the messages sent by B to A. Specifically, we will present to you 36 messages that were sent by B to A after B learned that there would be a Game 2. In all 36 instances, in Game 1, B promised to return a certain amount and then after A chose IN, B chose to transfer back to A an amount that was lower than the amount promised. We will call these participants “promise-breakers,” since they broke their promises in Game 1 by returning less than they promised they would.
Your job, and that of your partners in the matching game, is to make judgments about whether these written messages sent by promise-breakers after Game 1 and before Game 2 qualify as either of two definitions of “apologies.” You and your partners are trying to match your judgments about each of the messages. The more messages you match, the more you earn. Another way of thinking about this is as follows: You are trying to guess how your two partners will judge each message you all see and to make matching judgments yourself. The more accurate your guess about their judgments, the more money you can make.
We will provide you with participant-specific data on the Game 1 promise made, the Game 1 return made, and the subsequent message sent. Then, you and your partners will have to answer YES or NO to the following two
Does the message qualify as an apology according to apology definition 1?
Does the message qualify as an apology according to apology definition 2?
We cannot tell you how to play this matching game. Obviously, a lot of doing well depends on what your partners do. However, we recommend that, as you read each message, you and your partners pay attention to what you expect most people would interpret according to definition 1 and definition 2. We recommend that you DO NOT think too deeply about it. If you and your partners coordinate on the same intuitive approach, your initial interpretations should be most likely to pay off in this game. They are more likely to be similar for you and your partners than alternative interpretations produced by any deep thinking.
TWO DEFINITIONS OF APOLOGY THAT YOU WILL MAKE QUALIFYING JUDGEMENTS ABOUTApology definition 1: an explicit or implicit acknowledgment of offense which another has received Apology definition 2: an explicit or implicit acknowledgment of offense which another has received, along with remorse, regret, or sorrow stemming from acknowledgment of offense
You earn money in the matching game by making a judgment that agrees with the majority judgment of the group. In other words, if you match your judgment with at least one other partner – the two of you who have matched have formed a majority (i.e., at least 2 out of 3) and will be rewarded with $5 if the item you matched on is chosen for payment. If your judgment about a message does not match either of your partners’, you will not earn a money reward for that judgment.
Specifically, for each of the 36 messages you view, you will make YES or NO judgments in each of two columns that follow. In the first column you will circle either YES or NO, depending whether you think that the message content qualifies as an apology according to definition 1. In the second column you will circle either YES or NO, depending on whether you think that the message content qualifies as an apology according to definition 2.
Then for each individual message we will compare your judgment with judgments of the two other students. To calculate your earnings at the end of the experiment, we will randomly choose 3 out of 36 judgments in the first column and 3 out of 36 judgments in the second column for payment, using a bingo cage which contains balls numbered from 1 to 36. Since for each correct judgment you can earn $5, you can earn up to $30 for the matching game in addition to the $7 participation fee.
We will distribute to you a set of 36 Game 1 promise-breaker’s promise and return decisions and subsequent messages. In all, you will view 36 messages. Each message will be followed by two columns, each with YES or NO. In the first column you will circle either YES or NO, depending whether you think that the message content qualifies as an apology according to definition 1. In the second column you will circle either YES or NO, depending on whether you think that the message content qualifies as an apology according to definition 2.
Since we will choose three judgments in column 1 and three judgments in column 2, each potentially earning a reward of $5.00, you can earn up to $30 for the matching game. You additionally will earn $7 for showing up on time and participating.
Appendix C: Promises and Messages