Working Paper: CEPR ID: DP16988
Authors: Alexandra Avdeenko; Matthias Stelter
Abstract: In an increasingly data-driven world, data protection and the requirement of obtaining informed consent rapidly gain relevance. The intention is to protect data holders. Yet, is consent provided by data holders truly informed? In the context of empirical research, the requirement for informed consent can affect external validity and data quality of the evidence generated. Conducting a survey with 7,752 potential participants in rural Pakistan, we find that respondents are insufficiently informed about important aspects related to their consent. Experimentally changing the consent process, we find that showing an animated video has a negative impact on respondent’s understanding, but additionally engaging them in an interactive dialogue about the informational text significantly improves understanding. Even though we find effects on levels of understanding, we do not find meaningful changes in consent rates and non-response behavior indicating no adverse effects on the quality of the survey.
Keywords: ethics; survey methods; data collection; randomized control trial
JEL Codes: A13; C83; C93
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Standard consent process (D70) | Insufficiently informed about key aspects of consent (D82) |
| Animated video (Y10) | Decrease in understanding (D89) |
| Animated video + Interactive dialogue (Y10) | Improvement in understanding (O39) |
| Improved understanding (Y50) | No meaningful changes in consent rates (D19) |
| Improved understanding (Y50) | No adverse effect on response quality (C83) |