Interviewer Effects on Identity

How to cite

Hale, Henry; Onuch, Olga; Sasse, Gwendolyn; Kulyk, Volodymyr. 2020. "Interviewer Effects on Identity." AEA RCT Registry. AEARCTR-0005705. https://doi.org/10.1257/rct.5705-1.0

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

Henry Hale* , Olga Onuch , Gwendolyn Sasse , Volodymyr Kulyk

* Primary investigator

Registry ID
AEARCTR-0005705
DOI
https://doi.org/10.1257/rct.5705-1.0
First registered
Last updated
Status
In development
Trial dates
2020-04-13 – 2020-12-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/5705

Abstract

This study randomizes the assignment of interviewers with different ethnic backgrounds to respondents in a nationwide telephone poll of 2000 residents of Ukraine, a multi-ethnic country. This will enable us to test whether respondents systematically self-represent their own identity as well as political views often associated with identity in a survey setting differently when interviewed by interviewers who have different identity traits likely to be discernible through speech (over the phone). The relevant interviewer traits are language of childhood (Russian and Ukrainian) and region where one primarily lived prior to age 14. The study will explore heterogeneity of effects across regions of Ukraine and communities of varying degrees of diversity.

Intervention

All respondents in a nationally representative sample of 2000 adults in Ukraine will be randomly assigned an interviewer from KIIS’s call center, where interviewers have different identity backgrounds that are generally discernible through speech patterns. Assigned interviewers will vary on childhood language and region of primary residence through age 14. Our expectation is that interviewees will tend to report belonging to ethnic and linguistic identity categories in line with those of the interviewer. We also expect that interviewees will tend to report positions on attitude questions that are widely associated in Ukraine with the ethnic categories represented by the interviewers (we measure support for NATO, for compromise with Russia, and for Russian as a second state language).

Experimental design

The experiment (full text attached to this registry) is embedded in the regular “omnibus” survey of the highly reputable Kyiv International Institute of Sociology (KIIS). The survey is based on a stratified random sample with the strata being defined by the three-digit main operator prefixes and random generation of the rest of the number. Interviews will be conducted by phone using the CATI method.

Design details

Randomization method
Randomized assignment of interviewers to specific numbers is achieved by the randomization function in KIIS’s survey software.
Randomization unit
Individual
Clusters
0
Observations
2000 adult residents of Ukraine
Arms
50-70 interviewers in the call center will be randomly assigned to the 2000 respondents.
Primary outcomes
Our central outcomes of interest are self-reported identity (language chosen for the survey, nationality, native language, language of everyday life), attachment to ethnic group and country, and attitudes to NATO, compromise with Russia, and making Russian a second state language in Ukraine.

Cited by

No citation data available for this registration yet.

Keywords