Impact of additional e-Training on performance of research interviewers

How to cite

Chaparala, Sohini; Subramanian, Sneha; Haque, Mohammad Ashraful. 2021. "Impact of additional e-Training on performance of research interviewers." AEA RCT Registry. AEARCTR-0007446. https://doi.org/10.1257/rct.7446-1.0

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

Sohini Chaparala* , Sneha Subramanian , Mohammad Ashraful Haque

* Primary investigator

Registry ID
AEARCTR-0007446
DOI
https://doi.org/10.1257/rct.7446-1.0
First registered
Last updated
Status
On going
Trial dates
2021-04-01 – 2021-12-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/7446

Abstract

The confidence in evidence from experimental studies hinges on the quality of primary data collected through in-person interviews. Interviewing for rigorous research requires knowledge and certain soft skills. Interviewers or enumerators, who come from varying backgrounds are traditionally trained via classroom instruction, where a trainer delivers content, reinforcing understanding with role-plays and mock-interviews. Whereas e-Learning modes are touted to be better mediums for individualized learning of complex ideas, they are also cost-effective, scalable and have become indispensable in the post-pandemic world. We propose to study the impact of additional, learner-oriented e-training on the performance of research interviewers, on their knowledge and skills and on retention of this learning in the medium-term. We will offer intensive e-training to randomly selected interviewers, in addition to their project-specific training, and follow their knowledge and performance through a combination of objective and self-reported measures before, during and 6 months after the e-training. This experiment will contribute to the discussion on mitigation of measurement errors in burgeoning experimental research.

Intervention

The treated enumerators will remotely go through a set of training videos in Bangla, covering modules on scientific and ethical surveying, digital literacy, communication and teamwork etc. These videos cover training content in depth using multimedia and also include embedded quizzes to scaffold learning. The enumerators will engage with this material at their own pace within a 6-hour window. They will be compensated for their time spent on the training.

Experimental design

We propose a simple experimental design where 300 interviewers will be equally but randomly assigned, stratified by gender, education and experience, across the two groups: (a) Control Arm – this group will receive regular, synchronous training on projects, and (b) Treatment Arm – this group will receive regular, synchronous training on projects and additional e-training with embedded assessments. We will follow all interviewers through pre-intervention and post-intervention web surveys, while also collecting administrative data on their performance during the study.

Design details

Randomization method
The stratified random selection will be executed on the Stata software.
Randomization unit
The unit of randomization is the individual interviewer.
Clusters
No clusters are planned
Observations
300
Arms
150 treatment, 150 control
Primary outcomes
Average daily productivity on interviews; Average data error rates; Job-related knowledge; Soft skills measured by Big5
Secondary outcomes
Interviewer’s perceived competency; Interviewer's aspirations, expectations and job satisfaction

Cited by

No citation data available for this registration yet.

Keywords