Selective Memory of Goals as Reference Points

Last registered on December 08, 2025
View Trial History

Pre-Trial

Trial Information

General Information

Title
Selective Memory of Goals as Reference Points
RCT ID
AEARCTR-0015551
Initial registration date
March 26, 2025

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
April 03, 2025, 11:03 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated
December 08, 2025, 10:59 PM EST

Last updated is the most recent time when changes to the trial's registration were published.

Locations

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Primary Investigator

Name
Ao Wang
Affiliation
National University of Singapore
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Maximilian Mueller
PI Affiliation
Toulouse School of Economics
Contact Investigator
PI Name
Lony Bessagnet
PI Affiliation
Toulouse School of Economics
Contact Investigator

Additional Trial Information

Status
In development
Start date
2025-03-07
End date
2027-12-25
Keywords
Behavior, Education
Additional Keywords
JEL code(s)
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
The objective of this study is to understand the role of memory on the formation of educational goals regarding key metrics of academic performance, such as high school grade average. As the most important educational goals are often set for long-term purposes, their impact on educational effort or motivation depends on its persistence. On the other hand, however, goals have been recognized as a plausible candidate for reference points that bring sensation of gains and losses, and such sensation may induce students to manipulate them for the sake of wellbeing. This mechanism implies that long-term goals may be moderated by imperfect, and potentially biased recall of the original goals that one sets in the past. Armed with experimental variation, we seek to answer the following questions: (I) Whether recall of past goal is imperfect or biased, and how such imperfection or bias varies with one's current academic performance relative to one's original goal; (II) Whether incentives to enhance recall accuracy, and whether one's current relative academic performance moderates such effect; (III) Whether pre-announced incentives to remember the original goal enhances recall accuracy and educational performance.
External Link(s)

Registration Citation

Citation
Bessagnet, Lony, Maximilian Mueller and Ao Wang. 2025. "Selective Memory of Goals as Reference Points." AEA RCT Registry. December 08. https://doi.org/10.1257/rct.15551-1.1
Sponsors & Partners
Experimental Details

Interventions

Intervention(s)
As elaborated in the "Experimental Design" section, students are randomized into seven groups (Groups A through G) over the course of three surveys. Group G serves as the control group: while students in this group are asked in Surveys 2 and 3 to recall the goal they stated in Survey 1, these recall questions are not incentivized, and no advance notice is given in Survey 1.

In contrast, for the treatment groups (Groups A–F), the goal recall questions in Surveys 2 and/or 3 are incentivized. Additionally, some of these groups receive advance notice in Survey 1 about the upcoming recall questions.
Intervention Start Date
2025-03-07
Intervention End Date
2027-12-25

Primary Outcomes

Primary Outcomes (end points)
recalled goal, (relative) academic performance, stated wellbeing, stated wellbeing as a function of counterfactual academic performance [note: we plan to restrict our analysis of stated wellbeing questions to those who answer the wellbeing questions before the recall questions]
Primary Outcomes (explanation)
The stated wellbeing as a function of counterfactual academic performance can help us estimate the shape of reference dependence utility function. The sample restriction for the wellbeing questions is motivated by the fact that the items answered at the beginning of the survey—prior to any recall questions—are uncontaminated by subsequent survey content. This ensures that our analysis captures the direct effect of exam results, GPA, and information from previous surveys on subjective wellbeing, without interference from other questions in the instrument.

Secondary Outcomes

Secondary Outcomes (end points)
recall accuracy for the overall performance, response time for recall question, goal by subjects and their recall accuracy, perceived importance by subjects; recall accuracy of placebo questions (Stress during Exam Preparation and Attention Check)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
We plan to recruit 750 students who are about to take the Abitur. The final exams account for one third of their Abitur average, and they also function as the college entrance exams. These students would take three surveys. The key questions in the survey include the stated goal for the Abitur average, recollection of their stated goal and the recall accuracy, wellbeing as a function of academic performance, response time to key questions, as well as subject-specific performance, attitudes, and reported effort.

The experimental variations are implemented in the surveys. There are two major forms of variations to incentivize recalling the original goal. Ex-ante incentive refers to the kind of intervention where we inform students after they stated their goal of performance (Survey 1) that we will ask them about their original goal in specified waves in the future. Ex-post incentive refers to the kind of intervention where students are not given heads-up about the memory question in a specific wave, but questions that incentivize them to remember pop up unexpectedly in Survey 2 and 3. These two conditions are compared to None (No Incentives), which means that the question about goal recall is asked but not incentivized. In terms of these variations, we split the sample into seven groups (A, B, C, D, E, F, G)

Group A, B, C, D includes information in Survey 1 that informs students that the recall questions about their original goal will be asked in either Survey 2 or Survey 3. They will get an extra 5 euro if their answer is correct in the corresponding survey. The specific survey where the questions appear are listed below:
(A) (125 students) - Ex-ante for Survey 2, None for Survey 3,
(B) (125 students) - Ex-ante for Survey 2, Ex-post for Survey 3,
(C) (100 students) - None in Survey 2, Ex-ante for Survey 3,
(D) (100 students) - Ex-Post in Survey 2, Ex-ante for Survey 3,

In Group A and B, students are informed in Survey 1 that we will ask their original goal in an incentivized manner in Survey 2. In Group C and D, we will ask about their original goal in an incentivized manner in Survey 3. Note that for Ex-Post we would not inform students beforehand, although the incentivized goal recall question and its incentive will appear in the same form in the survey. None means that the question that ask students to recall also appear but would not be incentivized.

In Group E, F, G, students are not informed in Survey 1 that they might be questions that ask them recall their goal.
(E) (100 students) - Ex-post in Survey 2 + None in Survey 3
(F) (100 students) - None for Survey 2 + Ex-post for Survey 3
(G) (100 students) - None for Survey 2 + none for Survey 3

Note that not informing students in Survey 1 does not mean that the questions that ask students to recall does not appear. As we explained above, all Ex-post intervention includes the incentivized recall questions, so in Group E, the recall question in Survey 2 will actually be incentivized and the recall question in Survey 3 will not be incentivized, despite the plan that students will not be told about them in Survey 1.

In terms of the exact timing, Survey 1 is scheduled to take place right before Abitur exams take place. Survey 2 is scheduled to take place 4~6 months after the exams. Survey 3 is scheduled to take place 12~18 months after the exams.

Experimental Design Details
Not available
Randomization Method
Randomization by Qualtrics
Randomization Unit
Individual
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
About 90-100 schools
Sample size: planned number of observations
700 - 750 students (possibly in 2 waves). For later waves of surveys (Survey 2 and 3), we expect attrition of 10%.
Sample size (or number of clusters) by treatment arms
750 students.
Group A: 125 students
Group B: 125 students
Group C: 100 students
Group D: 100 students
Group E: 100 students
Group F: 100 students
Group G: 100 students
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
TSE Research Ethics Committee for Experimental Research
IRB Approval Date
2025-02-24
IRB Approval Number
N/A
Analysis Plan