Heterogeneity in meta-analysis refers to the variation in study outcomes between studies. I-square statistic describes the percentage of variation across studies while is calculated as the weighted sum of squared differences between individual study effects. Which is a classical measure of heterogeneity for meta-analysis (Cochran’s Q or Higgins I-Square)? and WHY?
Cochran's chi-squared test, also known as Cochran's Q test, is the classical measure of heterogeneity, if by "classical" you mean pioneering or first or historical. This technique has been known since 1954 (Cochran WG. The combination of estimates from different experiments. Biometrics. 1954;10:101–129.)
However, this measure is problematic when it comes to meta-analyses when the number of included studies is either relatively few or high, and has no intuitive interpretation. Higgins and Thompson I squared is among the relatively recent alternatives to the Q test that aims to address the low statistical power and the lack of intuitive interpretation of the Q test.
The following references explain this issue in greater detail:
1. https://www.statsdirect.com/help/meta_analysis/heterogeneity
2. Article Alternative Measures of Between-Study Heterogeneity in Meta-...
3. Article Comparison of four heterogeneity measures for meta-analysis
Chochrane'Q or Chi-square statistic is not reliable in case of few satudies.
I-statistic seems better option in such cases.
It is easy to interpret also.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis