Social Media Signalling

How to cite

Lowe, Matt; Angeli, Deivis. 2022. "Social Media Signalling." AEA RCT Registry. AEARCTR-0009359. https://doi.org/10.1257/rct.9359-2.1

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

Matt Lowe* , Deivis Angeli

* Primary investigator

Registry ID
AEARCTR-0009359
DOI
https://doi.org/10.1257/rct.9359-2.1
First registered
Last updated
Status
Completed
Trial dates
2022-05-12 – 2022-12-27
Countries
United States of America ()
Registry record
http://www.socialscienceregistry.org/trials/9359

Abstract

This trial studies social media signalling equilibria by exploring whether social media activity is predictive of offline behaviors. See pre-analysis plan (once available) for full details.

Intervention

See pre-analysis plan (once available) for full intervention details.

Experimental design

See pre-analysis plan (once available) for full experiment design details.

The pre-analysis plan relates to the audit study part of the paper. In addition, we are running a survey experiment (through CloudResearch's Connect platform) to complement the findings from the audit study.

In the survey experiment, we aim to recruit 1,000-1,500 American Twitter users to answer a survey on "social media use, well-being, and behavior" (recruitment is ongoing). At the end of the survey, we give them the chance to choose to donate some of their $5 survey incentive to the NAACP. The donation constitutes our private measure of racial attitudes. In the survey we also ask participants, in random order, (i) whether they want to retweet an NAACP tweet (we later track whether they retweeted or not), and (ii) for them to report privately how much they support the NAACP. We will then test for (i) whether those that retweet give more in NAACP donations than those that do not, and (ii) whether the informativeness of the retweet action is greater or less than the informativeness of the private measure of support.

Design details

Randomization method
Stratified randomization in STATA, full explanation in pre-analysis plan.
Randomization unit
Individual
Clusters
11k (more details in pre-analysis plan)
Observations
18,514 (more details in pre-analysis plan)
Arms
Core randomizations are 50:50
Primary outcomes
See pre-analysis plan (once available) for primary and secondary outcomes.

Cited by

No citation data available for this registration yet.

Keywords