Assessing heterogeneity in meta-analysis: Q statistic or I2 index?

3,932 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Linking Goal Progress and Subjective Well-Being: A Meta-analysis (2015, positive): “Additionally, we report I (Higgins and Thompson 2002) and the variance accounted for by artifacts (% VE; Hunter and Schmidt 2004) as a supplement to Q because first, in many situations the power of the Q test is very small (Harwell 1997), and second, I and % VE quantify the amount of heterogeneity a”
The Effect of Youth Diversion Programs on Recidivism (2013, methodology): “05) indicate that there are significant differences among studies (Huedo-Medina et al., 2006).”
Parent–Child Interaction Therapy for Child Disruptive Behaviour Disorders: A Meta-analysis (2016, neutral): “…statistically signiﬁcant and thus, the null hypothesis of homogeneity was not rejected, Q(11) = 8.85, p = .64; however, this test is underpowered (Huedo-Medina et al. 2006) and a random-effects model still remains appropriate because in the absence of heterogeneity this model would reveal the…”
How Effective are Instructional Explanations in Example-Based Learning? A Meta-Analytic Review (2010, background): “Note that the Q test usually has a relatively low statistical power when the number of studies is small (e.g., Huedo-Medina et al. 2006 ).”