Crowdsourcing the implicit association test: Limitations and best practices
Abstract
Although the use of crowdsourced online panels for behavioral data collection is commonplace in media and advertising research, only recently have software advancements made it possible for researchers to easily collect implicit measures online. Motivated by the recent decline in MTurk data quality and a dearth of literature examining the use of Implicit Association Tests with crowdsourced samples, we investigate cross-sectional data from eight IAT studies conducted using various samples (Mturk, online undergraduate students, and undergraduate behavioral labs). We document relative rates of participant inattention, non-naivety, and lack of motivation between crowdsourced and traditional samples and demonstrate the ramifications of these threats to the reliability and validity of IAT results. Finally, we build on these insights to outline best practices for crowdsourcing implicit measures in advertising and media research.
Study specs
- Authors
- S Connors,K Spangenberg,AW Perkins
- Discipline
- Behavioral Science
- Year
- 2020
- Human Data Platform
- Prolific
- Source
- View Source DOI Google Scholar
Peer Review & Critical Discussion
Potential Selection Bias in 2023 Cohort
The participant pool shows a concerning overrepresentation of users from high-income demographics. Looking at Table 3, we can see that 78% of respondents had annual incomes above $75k, which significantly limits the generalizability of these findings to broader populations.
Non-naive Participants Issue
I've noticed a methodological concern regarding participant naivety. Given that Prolific users often complete multiple studies, there's a real risk that participants had prior exposure to similar experimental paradigms, which could confound the results.
RLHF Applicability to This Study Design
The implications for RLHF training pipelines are understated. If we accept the authors' conclusions about preference stability, this has direct consequences for how we should structure reward model training. The temporal decay effect described in Section 4.2 is particularly relevant.
Verify your expertise to join discussion
Create an account and verify your credentials to participate in peer discussions.