New Experiment Unveils Patterns in Old Social Science Data

Research from MIT Sloan School of Management demonstrates a new way of designing social science experiments to uncover hidden patterns.

Apr. 10, 2026 at 4:36am

Got story updates? Submit your updates here. ›

A bold, highly structured abstract painting in muted earth tones, featuring sweeping geometric arcs, concentric circles, and precise botanical spirals, conceptually representing the complex interactions that shape social and behavioral outcomes.An abstract visualization of the intricate interplay between factors that determine the impact of punishment on collective welfare in social dilemmas.Philadelphia Today

Researchers from MIT Sloan School of Management, Cornell University, and the University of Pennsylvania have developed a new approach called 'integrative experiment design' that can uncover complex patterns in social science data that were previously invisible. By systematically varying multiple factors across hundreds of experimental conditions, the team was able to model how punishment affects collective welfare in public goods games, finding that the impact can swing from harmful to helpful depending on factors like communication and contribution framing.

Why it matters

Social and behavioral sciences have long struggled to integrate findings across many isolated studies, often generating a long list of factors that matter without a clear picture of how they fit together. The integrative experiment design approach addresses this by making integration a core part of the research process from the start, allowing researchers to model complex interactions and make predictions about outcomes in new conditions.

The details

The researchers applied their integrative approach to the longstanding question of when punishment helps or harms collective welfare in public goods games. They systematically varied 14 design parameters - including group size, game length, communication, and contribution framing - across 360 experimental conditions involving thousands of participants. They found that punishment's effect can swing from substantially harmful to substantially helpful depending on the specific combination of factors, with communication emerging as the most important predictor, followed by contribution framing.

  • The research paper was published in the journal Science on April 10, 2026.
  • The integrative experiment design framework was first proposed by the researchers in Behavioral and Brain Sciences in 2024.

The players

Abdullah Almaatouq

An associate professor at the MIT Sloan School of Management and a co-author of the research paper.

Mohammed Alsobay

A recent Ph.D. graduate in Information Technology from MIT Sloan and a co-author of the research paper.

David G. Rand

A professor at Cornell University and a co-author of the research paper.

Duncan J. Watts

A professor at the University of Pennsylvania and a co-author of the research paper.

Empirica

An open-source platform for running large-scale social science experiments, developed by the research team to enable their integrative approach.

Got photos? Submit your photos here. ›

What they’re saying

“Research programs can be seemingly productive for a long time — generating study after study — and still not know what they know. The integration of findings across studies is assumed to happen through the publishing process, but it often doesn't, because the studies weren't designed to be put together in the first place.”

— Abdullah Almaatouq, Associate Professor, MIT Sloan School of Management

“Because the outcome depends on many factors that interact in complex ways, more than two decades of research and thousands of papers have not been able to pin down the conditions under which punishment improves social welfare. Examining punishment in the context of public goods games is a beachhead demonstration for our integrative framework, getting us closer to more context-sensitive and practically useful behavioral science.”

— Mohammed Alsobay, Recent Ph.D. Graduate, MIT Sloan

“Our approach enables us to predict what will happen in experimental conditions we haven't observed before, comparing favorably to both laypeople, and researchers asked to make the same predictions. The patterns are complex, but they are stable and learnable.”

— Mohammed Alsobay, Recent Ph.D. Graduate, MIT Sloan

What’s next

The researchers plan to apply the integrative experiment design approach to other domains in social and behavioral science to further demonstrate its ability to uncover hidden patterns and make context-sensitive predictions.

The takeaway

This research showcases a powerful new approach to social science experiments that moves beyond studying factors in isolation and instead models the complex interactions that shape real-world outcomes. By systematically varying multiple factors, the team was able to uncover nuanced insights about when punishment helps or harms collective welfare - insights that could inform more effective policies and interventions.