Targeting Goals

How to cite

Talantceva, Marina; Sliwka, Dirk; Zimmermann, Tom; Opitz, Saskia; Vogelsang, Timo. 2024. "Targeting Goals." AEA RCT Registry. AEARCTR-0013088. https://doi.org/10.1257/rct.13088-1.0

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

Marina Talantceva* , Dirk Sliwka , Tom Zimmermann , Saskia Opitz , Timo Vogelsang

* Primary investigator

Registry ID
AEARCTR-0013088
DOI
https://doi.org/10.1257/rct.13088-1.0
First registered
Last updated
Status
In development
Trial dates
2024-02-26 – 2024-07-31
Countries
Private
JEL codes
C21, C93, M50
Registry record
http://www.socialscienceregistry.org/trials/13088

Abstract

A large literature in economics, management and psychology has shown that performance is affected by goals set by employers and that the relationship between the specific goal size and performance is inversely U-shaped. In this study, we ask the question whether and to what extent the goals can effectively be targeted to individual people based on their prior performance and, importantly, their personal traits (preferences, personality traits and beliefs).

Intervention

Participants first fill out a survey on demographics, preferences, and personality traits and work on task. In a second part subjects are again invited to work on a real effort task. This time, a specific performance goal is randomly assigned to each subject and subjects receive a bonus when reaching the goal. This is a pre-registration of a pilot experiment.

Experimental design

The Prolific experiment will consist of two parts. In the first part, participants have to complete a survey, where information about participants’ demographics (age, gender, education, race), risk preferences, ambition, level of self-esteem, beliefs about their own performance, and Big-5 Personality traits will be collected. In addition, an open question (“If you think about yourself: Are you a person who is motivated by challenges or do too challenging tasks rather demotivate you? Please explain also why this is the case in your view.”) will be asked. After that, participants complete a real-effort task for 10 minutes with an opportunity to train themselves for 30 seconds before. They are not incentivized and receive only a participation fee for completing this part of the experiment.

The real-effort task follows DellaVigna and Pope (2018): Participants have to alternately press 'a' and 'b' buttons on their keyboards. For each correct sequence pressed, they get 1 point. All participants receive information about the number of gained points on their screen.
We invite participants again after approximately one week to participate in the second part of the experiment. We will only invite participants who have completed all the sections from the first part of the study. In the second part, they again have to do the same real-effort task for 10 minutes. However, before starting with the task, each of them is assigned with a goal, which is a number of points that they have to gain to receive a financial bonus. After finishing the real-effort task, participants fill out a short post-survey consisting of questions about their enjoyment of the task. At the end of the experiment, workers will be redirected to Prolific, and will receive the participation fees and bonuses (for those who reached the goal).

Design details

Randomization method
The goals are randomly assigned. The assignment follows a probability distribution is constructed as follows: We start with the performance distribution obtained in a study that used the same real-effort task (Opitz, Sliwka, Vogelsang, & Zimmermann, 2022). The range of the values from the previous study is split to quintiles. Participants are first randomly assigned to each quintile, and then receive a random target value drawn from the range of performance outcomes in the respective quintile. That is done for the purpose of having more observations in the range of performance likely to be achieved and less observations for extreme values (where likely either all subjects reach the target or no one is able to do).
Randomization unit
Individual Subject
Clusters
We invite 300 individual subjects to participate in the study as it is a pilot.
Observations
We invite 300 individual subjects to participate in the study as it is a pilot.
Arms
All the subjects (300 participants) receive a goal treatment.
Primary outcomes
The primary outcome variable is the number of points gained in the real-effort task. The aim of this study is to estimate Conditional Average Treatment Effects of variations in goal size based on prior performance and personal characteristics of the participants. We will use this to estimate (i) how the size of the performance-maximizing goal depends on prior performance and a person’s preferences and personality and (ii) whether performance can be increased by targeting goals to individuals based on their characteristics.
Secondary outcomes
Enjoyment of the task.

Cited by

No citation data available for this registration yet.

Keywords