IPV Measurement: Application and Validation of the List Experiment

How to cite

Mughogho, Winnie; Owsley, Nicholas; Yanganaram, Dhwani. 2021. "IPV Measurement: Application and Validation of the List Experiment." AEA RCT Registry. AEARCTR-0008558. https://doi.org/10.1257/rct.8558-1.0

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

Winnie Mughogho* , Nicholas Owsley , Dhwani Yanganaram

* Primary investigator

Registry ID
AEARCTR-0008558
DOI
https://doi.org/10.1257/rct.8558-1.0
First registered
Last updated
Status
In development
Trial dates
2022-01-15 – 2022-03-15
Countries
Kenya ()
Registry record
http://www.socialscienceregistry.org/trials/8558

Abstract

Obtaining accurate information on behaviors and attitudes considered sensitive has posed a challenge for social science researchers. Directly asking sensitive questions can lead to misreporting or nonresponses as respondents may misrepresent their true responses for fear of repercussions, to conform to social norms, or to avoid shame. These inaccurate responses introduce systematic measurement error and bias to the estimates of sensitive behaviours and subsequent inferences. To mitigate this bias, indirect response (IR) techniques, such as the List Experiment (LE), have been used in surveys to better identify estimates of sensitive behavior. We will apply the List Experiment to measure and characterize inconsistencies in reporting of Intimate Partner Violence in a Kenyan sample. We will further test the validity of the technique applying novel methods in the context of the Global South though generalizable to other contexts.

Experimental design

In order to obtain prevalence levels of IPV within our sample, we will apply the Double List Experiment and also obtain information from direct report. We will randomly assign respondents two groups, where each group will alternate as treatment and control for different lists.

Design details

Randomization method
Computer
Randomization unit
Individual
Clusters
0
Observations
1000 respondents
Arms
500 respondents per mode of data collection, and 250 per arm under each mode
Primary outcomes
Inconsistency of IPV reporting; that is the difference between estimates from direct method and indirect method.
Secondary outcomes
- Inconsistency in reporting by subgroup - Inconsistency in reporting across modes - Validity of measures

Cited by

No citation data available for this registration yet.

Keywords