Strategic Reasoning and Perspective Taking

How to cite

Falco, Paolo; Fafchamps, Marcel; Chegere, Martin. 2023. "Strategic Reasoning and Perspective Taking." AEA RCT Registry. AEARCTR-0010717. https://doi.org/10.1257/rct.10717-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Paolo Falco* , Marcel Fafchamps , Martin Chegere

* Primary investigator

Registry ID
AEARCTR-0010717
DOI
https://doi.org/10.1257/rct.10717-1.1
First registered
Last updated
Status
In development
Trial dates
2023-01-09 – 2023-03-09
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/10717

Abstract

We will conduct an experiment to investigate how strategic decisions are affected by information about the opponent’s characteristics. Subjects will play a series of 22 guessing tasks where they will have to guess a number as close as possible to an anonymous opponent’s guess times a multiplier. By means of randomized treatments, we will study how their decisions are affected by revealing specific information about the opponent. In doing so, we will leverage a growing literature on perspective-taking, which suggests that being able to take someone else’s perspective may guide decision-making in strategic settings.

Experimental design

We will conduct an experiment with university students to investigate how strategic decisions are affected by information about the opponent. Subjects will play a series of 22 guessing tasks where they will have to guess a number as close as possible to an anonymous opponent’s guess times a multiplier. The parameters of the game and the opponent will change from task to task.

In 6 out of the 22 tasks (Control), we will reveal no information about the opponent.

In the remaining 16 tasks (Treated), we will reveal key characteristics of the opponent: gender (MALE/FEMALE), field of study (STEM/NON-STEM subject), and Grade Point Average (TOP GRADE/LOW GRADE). Combining these traits produces 8 opponent profiles. We will have two treatment arms and use each profile twice (once per treatment arm). In the first treatment arm (“INFO ONLY”), we will only reveal the traits listed above (subjects will read the information on their screens). In the second treatment arm (“INFO + VIDEO”), we will reveal the traits listed above and, in addition, we will show a video of the opponent sharing some "fun facts" about himself/herself (e.g., favorite food, favorite activity, etc.) The combination of 8 profiles and two treatment arms results in 16 treatments. Each treatment will be randomly assigned to one of the 16 treated tasks and the order of these tasks will differ between subjects. Conversely, the same four control tasks will be the first ones to be played by all subjects, and the remaining two control tasks will be the last ones to be played by all subjects (after all the treated tasks have been completed).

For 60 subjects out of 200, in the second treatment arm, we will only show the video and we will not reveal the opponent’s characteristics (“VIDEO ONLY”).

Subjects will also play a dictator game with two randomly chosen opponents over the course of the experiment.

Design details

Randomization method
Randomization will be conducted by means of a computer program.
Randomization unit
Guessing task. Each individual will play 22 guessing tasks. We will randomize the treatments across tasks within an individual and the order of the tasks will differ between individuals. For 140 individuals, the videos will be accompanied by information about the opponent (INFO + VIDEO). For 60 individuals, the videos will not be accompanied by information about the opponent (VIDEO ONLY).
Clusters
N/A
Observations
4,400 guessing tasks (22 guessing tasks per individual x 200 individuals)
Arms
Control: 1,200 (6 guessing tasks x 200 individuals); Treatment 1 - 16 with INFO ONLY and INFO + VIDEO: 140 (1 guessing task x 140 individuals) per treatment; Treatment 1 - 16 with INFO ONLY and VIDEO ONLY: 60 (1 guessing task x 60 individuals) per treatment.
Primary outcomes
Guesses; Degree of rationality implied by guesses; Beliefs about opponents' guesses; Degree of opponents' rationality implied by beliefs about opponents' guesses; Choices in a dictator game. In parts of the analysis, we will focus on guesses that can be rationalized (i.e., we will drop guesses that appear to be random, confused, or cannot be rationalized).

Cited by

No citation data available for this registration yet.

Keywords