The "Naturalistic Free Recall" dataset: How Prolific supported pioneering memory research

Challenge 

How do humans remember and retell stories? Researchers at Yale and NYU wanted to understand how a fundamental but understudied aspect of the human memory system works. Traditional memory studies relied on simple list-learning, but the team were keen to explore how people recall complex narratives in natural settings.

To do this, they needed to:

• Recruit hundreds of participants willing to listen to and verbally recall lengthy narratives
• Ensure high-quality audio recordings with professional transcription
• Collect comprehensive demographic data across diverse age groups
• Maintain research integrity through a complex, multi-stage protocol

Solution 

The researchers needed a solution that could handle sophisticated data collection while maintaining rigorous quality standards. Prolific helped deliver part of this ambitious study.

Quality-focused recruitment

Prolific was the perfect fit for the study, helping to recruit 62 participants from a diverse range of backgrounds, with an average age of 39.77 years. 

Our pre-screening tools ensured participants met strict research criteria, and the study design included clear audio instructions followed by on-screen text guidance, plus visual indicators to help participants track their recording times.

Reliable data collection

Participants completed multiple audio recordings throughout the study, each lasting over four minutes. The study integrated with PsychoPy for experiment delivery and used professional transcription services to ensure data quality, helping maintain consistency across all submissions.

Execution 

The study followed a clear protocol. Participants listened to two of four possible stories, each eight to 13 minutes long, from The Moth Radio Hour podcast series and an audiobook. After each story, they provided detailed verbal recalls and completed questionnaires about memory strategies and task engagement. Detailed demographic data was also collected to better  understand individual differences.

Results

The study revealed important insights about how humans remember and retell stories:

• Initial events in narratives showed significantly higher probability of being recalled first across all stories
• Participants typically recalled events in the same order they were presented, showing strong temporal relationships
• The study validated these classic memory patterns in natural storytelling for the first time
• Some narrative events were consistently remembered well across participants while others were frequently forgotten
• Story elements with strong semantic connections to other parts of the narrative were more likely to be recalled

Conclusion 

Conducted by researchers from Yale, NYU, Max Planck Institute, and Intel Labs, the study demonstrates how Prolific can support sophisticated research protocols while maintaining high data quality standards. The success of this complex study highlights our ability to facilitate advanced behavioral research requiring careful participant selection and reliable data collection.

For researchers tackling ambitious studies requiring detailed participant responses, Prolific offers:

• Access to engaged, diverse participants
• Support for complex research protocols
• Robust quality control measures
• Integration with research tools

Citation: Raccah, O., Chen, P., Gureckis, T.M., Poeppel, D., & Vo, V.A. (2024). The "Naturalistic Free Recall" dataset: four stories, hundreds of participants, and high-fidelity transcriptions. Scientific Data, 11:1317.

Research institutions: Yale University, New York University, NYU & Max Planck Institute, Ernst Strüngmann Institute for Neuroscience, Intel Labs