Budgeting, Feedback, and Spending Decisions - Study Extension and Replication

How to cite

Beshears, John; Gillis, Talia. 2015. "Budgeting, Feedback, and Spending Decisions - Study Extension and Replication." AEA RCT Registry. AEARCTR-0000932. https://doi.org/10.1257/rct.932-7.0

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

John Beshears* , Talia Gillis

* Primary investigator

Registry ID
AEARCTR-0000932
DOI
https://doi.org/10.1257/rct.932-7.0
First registered
Last updated
Status
Withdrawn
Trial dates
2015-11-01 – 2016-01-01
Countries
Brazil ()
Registry record
http://www.socialscienceregistry.org/trials/932

Abstract

This trial extends the trial period for "Budgeting, Feedback, and Spending Decisions" (AEARCTR-0000889) and launches a replication of the original trial.

As of the date of this registration (October 30, 2015), the investigators have not received data regarding the spending decisions of individuals in the original study (AEARCTR-0000889). The partner company that is implementing the trial, however, has been tracking user engagement as a result of the trial (the frequency with which individuals in the study use the mobile app). The partner company has been pleased with the user engagement resulting from the trial and is expanding the research in two ways. First, individuals who are in the original trial will continue to receive the treatment to which they were originally randomly assigned beyond October. Second, additional individuals who are not part of the original trial will be randomized into a replication of the original trial.

The abstract from the original trial is pasted here: Working with the developers of a mobile phone app that helps individuals track their spending, we study whether giving users regular feedback (via push notifications) on their progress relative to their budgets influences spending on bank cards and credit cards. Users are randomly assigned to receive progress updates relative to narrow budget goals or relative to broad budget goals. A third group is randomly assigned not to receive progress updates.

Experimental design

The experimental design for the original trial (AEARCTR-0000889) is unchanged.

The experimental design for the replication study is the same as the experimental design for the original trial and includes three groups:
1) Broad budget goal - users receive push notifications providing feedback about spending relative to a broad budget goal
2) Narrow budget goal - users receive push notifications providing feedback about spending relative to a narrow budget goal
3) Control - users do not receive push notifications providing feedback on spending relative to budget goals

Design details

Randomization method
Randomization for the original trial (AEARCTR-0000889) has already taken place. Randomization for the replication study is done in an office by a computer.
Randomization unit
Randomization for the original trial (AEARCTR-0000889) has already taken place. The unit of randomization for the replication study is the individual user.
Clusters
Randomization for the original trial (AEARCTR-0000889) has already taken place. The sample for the replication study will include approximately 27000 users (randomization is at the user level).
Observations
Randomization for the original trial (AEARCTR-0000889) has already taken place. The sample for the replication study will include approximately 27000 users (randomization is at the user level).
Arms
Randomization for the original trial (AEARCTR-0000889) has already taken place. In the replication study, the sample sizes by treatment arm are approximately: ~12000 in each of two treatment groups and ~3000 in control group
Primary outcomes
The end points for the original trial (AEARCTR-0000889) are unchanged. For the replication study, the text here is pasted from the original trial (AEARCTR-0000889) but updates the dates for the replication study: Our key outcome variable will be bank card and credit card spending during November 2015 in the key category (shopping, bars/restaurants, or supermarket) with the largest budget. Because the distribution of this variable will be highly skewed, we will focus on the logarithm of the variable and also plan to winsorize it (at the 99th percentile of its distribution, unless the distribution is so extreme that the 95th percentile is more appropriate). In order to increase statistical precision, we also plan to look at changes in the variable relative to October 2015 (it may not be possible, however, to obtain October 2015 data). As secondary analyses, we plan to conduct quantile regressions to study differences in the full distribution of the key outcome variable across treatment groups. We also plan to examine secondary outcome variables such as an indicator for whether or not the goal was achieved and actual spending as a proportion of budgeted spending. Finally, tertiary analyses will examine overall spending (as opposed to spending only within the key category with the largest budget).

Cited by

No citation data available for this registration yet.

Keywords