Baseline testing – a widespread but difficult to treat disease

I once reviewed tables for a randomized study and noticed several comments about testing the baseline characteristics. The commenters were arguing which test would be best to test for the differences between the 2 randomized groups at baseline.

This made me first angry about the wasted time and then curious about the reasons, statisticians still do this.

In today’s episode, Benjamin and I discuss some backgrounds for baseline testing in randomized studies. We also recommend these further readings:

  • Baseline comparisons and covariate fishing: Bad statistical habits we should have broken yesterday, by Stefan L. K. Gruijters
  • Testing for baseline balance in clinical trials, by Stephen Senn

References:

http://statisticalmisses.nl

http://statisticalmisses.nl/index.php/faq/84-why-are-significance-tests-of-baseline-differences-a-very-bad-idea

Listen to this episode and become an effective statistician!

Scroll to Top