Timeline for Can effect sizes (SES) be compared when the underlying null distributions are different in a meta-analysis?
Current License: CC BY-SA 4.0
2 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 11, 2018 at 15:34 | comment | added | Mike Lavender | I might not be understanding your answer, but I am not assuming there is an error. On the contrary, I am assuming the model is correct and wondering if a fair interpretation is that because each "study" is significantly different, massively as you state, if in an applied sense I should be comparing the studies to each other. That is, each study is measuring something (some effect) but not the same effect. | |
| Sep 9, 2018 at 13:33 | history | answered | mdewey | CC BY-SA 4.0 |