Zee score, boss, zee score!

This one's from The Washington Monthly. There a new meta-study out about political science papers. They looked at the z-scores of results in many papers. These are measures of how significant results are. The paper says that a z-score of 1.96 is the boundary between significance and insignificance. They found that there are three times more papers with z-scores slightly above 1.96 than papers with z-scores slightly below 1.96. Meaning that there are lots of political scientists fudging their studies to make the results seem significant.


Post a Comment

<< Home