Meeting Register Page

TI Methods Speaker Series: The I-squared statistic does not tell us how much the effect-size varies in meta-analysis (Michael Borenstein)
The vast majority of meta-analyses published in medical journals use the I-squared statistic to quantify heterogeneity. While this use of I-squared is ubiquitous, it is nevertheless incorrect. The I-squared statistic does not tell us how much the effect size varies. An I-squared value of 25% may reflect trivial heterogeneity but may also reflect substantial heterogeneity. Conversely, an I-squared value of 75% may reflect substantial heterogeneity but may also reflect trivial heterogeneity. The statistic that does tell us how much the effect size varies is the prediction interval.  This statistic provides the information that we need to make informed decisions, and that many researchers mistakenly believe is being provided by I-squared.

Nov 30, 2022 12:00 PM in Vancouver

UBC Therapeutics Initiative: Better prescribing. Better health.
Meeting is over, you can not register now. If you have any questions, please contact Meeting host: Ciprian Jauca.