Managing Identity

How to cite

Coffman, Katherine; Aina, Chiara. 2026. "Managing Identity." AEA RCT Registry. AEARCTR-0018544. https://doi.org/10.1257/rct.18544-1.0

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

Katherine Coffman* , Chiara Aina

* Primary investigator

Registry ID
AEARCTR-0018544
DOI
https://doi.org/10.1257/rct.18544-1.0
First registered
Last updated
Status
In development
Trial dates
2026-05-04 – 2026-05-22
Countries
United States of America ()
JEL codes
J16, D91
Registry record
http://www.socialscienceregistry.org/trials/18544

Abstract

We are studying the “self-narratives” people hold: how they view themselves in terms of what they are good at and what is important to them, with a focus on how people manage these self-narratives in response to new information. We plan to collect both open-ended and more structured measures of individuals’ self-narratives by asking participants how they view different self-aspects of their identity: their intellectual self-aspect, their relational self-aspect, and their creative self-aspect. We will explore how these self-narratives correlate with objective measures of skill and individuals’ beliefs about their skills, as well as whether these self-narratives vary by gender. We will then expose some participants to feedback on their intellectual skills and use exogenous variation in the feedback to understand how individuals manage their self-narratives in response to new information. Does negative feedback in the intellectual self-aspect lead to less identification with that self-aspect, and more identification with other self-aspects?

Intervention

Participants will complete a single 20-minute survey.

In Part 1, they will answer questions related to how they see themselves, including an open-ended, audio-recorded question and more structured questions that ask them how they would describe themselves.

In Part 2, participants will complete a skills test in each of the three self-aspect domains (intellectual, relational, creative). Each test consists of 12 multiple-choice questions with 30- seconds allotted for each question. After, we will ask them questions about how they believed they performed both in absolute and relative terms.

Participants are then randomly assigned to one of three treatments:
No Feedback: Participants receive no feedback on their performance on the intellectual test.
Partial Easy: Participants are told how they performed on 6 of the 12 intellectual test questions, chosen ex-ante to be easier questions on average.
Partial Hard: Participants are told how they performed on 6 of the 12 intellectual test questions, chosen ex-ante to be harder questions on average.

In Part 3, participants answer more detailed questions related to how they see themselves across the different self-aspects. They also provide updated beliefs about how they performed on the tests from Part 2.

In Part 4, they will answer demographic questions.

Experimental design

Participants will complete a single 20-minute survey.

In Part 1, they will answer questions related to how they see themselves, including an open-ended, audio-recorded question and more structured questions that ask them how they would describe themselves.

In Part 2, participants will complete a skills test in each of the three self-aspect domains (intellectual, relational, creative). Each test consists of 12 multiple-choice questions with 30- seconds allotted for each question. After, we will ask them questions about how they believed they performed both in absolute and relative terms.

Participants are then randomly assigned to one of three treatments:
No Feedback: Participants receive no feedback on their performance on the intellectual test.
Partial Easy: Participants are told how they performed on 6 of the 12 intellectual test questions, chosen ex-ante to be easier questions on average.
Partial Hard: Participants are told how they performed on 6 of the 12 intellectual test questions, chosen ex-ante to be harder questions on average.

In Part 3, participants answer more detailed questions related to how they see themselves across the different self-aspects. They also provide updated beliefs about how they performed on the tests from Part 2.

In Part 4, they will answer demographic questions.

See uploaded document for experimental materials.

Design details

Randomization method
Randomization will be done by the experimental program (Qualtrics). We will stratify based upon the initial ranking of self-aspects provided (two groups: ranked intellectual first, or ranked intellectual second/third).
Randomization unit
Individual level
Clusters
1800 participants
Observations
1800 participants
Arms
600 in each of three treatments
Primary outcomes
Our primary measures aim to capture self-narratives across three self-aspects (SA) after treatment: intellectual, relational, and creative. We do this with a battery of key questions: SA RANK – This refers to the forced ranking an individual gives to the SA after treatment, where they are asked to rank the self-aspect that feels most important to them as number 1. SA SHARE – This refers to the zero-sum share that an individual gives to the SA after treatment, where they are asked, relative to each other, how important each SA is to what makes you, “you,” today. SA WORDS – This refers to the number of words from the SA category that an individual chooses in the word selection task after treatment. In this task, individuals are asked to select from a list of 24 different words (8 from each SA) all words that they would choose to describe themselves. SA BUBBLE – Individuals are asked after treatment how big a part each SA is to their overall identity on 1 - 7 scale. SA IDEAL TIME – Individuals are asked after treatment how they would ideally allocate their leisure time in a typical week across the considered SAs. In addition to these measures of identity, we are also interested in the beliefs individuals hold about their abilities in each SA post-treatment. SA Belief of Absolute Performance – Guess of their 12-question score on the multiple-choice SA test, post-treatment. SA Belief of Broader Skills – Self-assessment of their broader skills in that SA on a 1 - 7 scale where 1 is extremely weak and 7 is extremely strong.
Secondary outcomes
In addition to our primary self-narrative measures, we have a few additional measures that also aim to capture identity but in more nuanced, naturalistic ways that may be less susceptible to demand effects. In the first part of the study, we ask individuals to describe themselves in an open-ended question. We will use an LLM to analyze the extent to which these descriptions seem to reflect intellectual, relational, and creative aspects of their identity. Later in the study, we ask individuals to choose from a list of 6 possible books (2 for each SA) which book they would be most interested in. Because we randomly select some participants to indeed receive the e-book format of the book they choose, this is an incentivized measure of identity. We plan to see how these two measures correlate with our other measures of identity.

Cited by

No citation data available for this registration yet.

Keywords