Do Subtle Linguistic Interventions Priming a Social Identity as a Voter Have Outsized Effects on Voter Turnout? Evidence from a New Replication Experiment

ISPS Data Archive: Terms of Use

By using, contributing, and/or downloading files associated with scholarly studies available on the ISPS Data Archive, you agree to these terms and conditions. Please read the ISPS Data Archive Terms of Use.

Suggested citation: 

Alan Gerber, Greg Huber, and Al Fang (2020). Replication Materials for, ‘Do Subtle Linguistic Interventions Priming a Social Identity as a Voter Have Outsized Effects on Voter Turnout? Evidence from a New Replication Experiment,’ ISPS Data Archive.


Gregory Huber, Al Fang, Alan S. Gerber

Research design: 
Data type: 
Data source(s): 


Data source information: 


Field Date: 
2015-10-31 - 2015-11-21
Location details: 
Kentucky; Mississippi; Houston; Louisiana
Unit of observation: 
Sample size: 
Respondents recruited for the SSI sample were admitted into the experiment if their reported five-digit zip code for their place of residence is contained in a master set of zip codes associated with Kentucky, Louisiana, Mississippi, or Houston; if they reported being registered to vote at the time of the survey; if they are at least 18 years old, which is derived from their reported birth year; and if they provided a valid first and last name. Respondents recruited for the YouGov sample were admitted into the experiment if they were identified by YouGov as a registered voter 18 years of age or older and if they reported living in a zip code contained in the master zip code list.
Randomization procedure: 
Subjects are assigned to the noun condition or to the verb condition with equal probability using a simple random assignment procedure that each vendor implemented separately. Across both samples, 1,579 (51.3%) were assigned to the noun condition, and 1,499 (48.7%) were assigned to the verb condition. Because the randomization procedure occurred by vendor, we conduct a randomization check separately for each sample by regressing treatment assignment on a vector of pretreatment covariates and test the null hypothesis that these covariates are not jointly prognostic of treatment assignment. We fail to reject the null hypothesis for each sample and infer that the randomization procedures are valid.
Respondents admitted into the experiment as a subject are immediately randomly assigned to receive either a 10-item questionnaire using noun wording (“voter”) or a 10-item questionnaire using the verb wording that refers to the act of voting as a behavior (“voting/to vote”).
Treatment administration: 
Web delivered
Outcome measures: 
turnout in the 2015 general election
Archive date: 
Terms of use: 

ISPS Data Archive: Terms of Use.

Area of study: 
Data file number Description File format Sizesort descending File url
D156F10.1 Supplementary materials .pdf 161073 Download file
D156F06 Program file .R 2412 Download file
D156F02 Dataset .dta 242707 Download file
D156F01 Readme file .txt 2968 Download file
D156F11 Metadata (DDI 3.2) .xml 299393 Download file
D156F03 Program file .do 30764 Download file
D156F08 Output file .csv 362 Download file
D156F07 Output file .log 625481 Download file
D156F04 Program file .R 6696 Download file
D156F12 ISPS Data Archive Report .pdf 745369 Download file
D156F09 Output file .csv 747 Download file
D156F05 Output file .pdf 8025 Download file