Data and stories
1 min read · April 11, 2025
New Power Labs
A study by Silberzahn et al. (2018) found that analysts arrived at different conclusions even with the same dataset.
In this research, 29 teams of analysts were given the same dataset and asked to investigate whether soccer referees were more likely to give red cards to dark-skin-toned players than to light-skin-toned players. The results varied:
20 teams found a statistically significant relationship.
9 teams did not.
This divergence wasn’t due to errors, but rather to legitimate differences in:
Statistical techniques
Decisions about which variables to control for
How to handle outliers or missing data
This variation highlights how analytical choices—statistical methods, covariate selection, and data handling—shape outcomes. It underscores the need not just for reliable and representative datasets but also for transparency in data processing and analysis.
The way we handle data influences the stories it tells, and the gaps and opportunities we uncover.
Narinder
New Power Labs
Like what you’re reading? Subscribe to get weekly Equity Shots in your inbox.
