Table 3.
Comparisons of models for testing effect‐size differences between languages, with and without language as a fixed factor
| Meta‐analysis | Fixed factor(s) | Random factor(s) | Comparison of models with and without language as the fixed factor | ||
|---|---|---|---|---|---|
| F | x 2 | p | |||
| Rice‐field meta‐analysis | Language + Taxa | 0.26 | .62 | ||
| Language + Landscape | 0.14 | .71 | |||
| Language | Taxa | 1.97 | .16 | ||
| Language | Landscape | 0.58 | .45 | ||
| Language | Taxa + Landscape | 0.17 | .68 | ||
| Leaf life span meta‐analysis | Language + Measurement condition | 12.64 | .0005 | ||
| Language + Plant family | 12.76 | .0005 | |||
| Language | Measurement condition | 10.68 | .001 | ||
| Language | Plant family | 13.59 | .0002 | ||
| Language | Study country | 16.00 | <.0001 | ||
| Language | Measurement condition + Plant family + Study country | 16.10 | <.0001 | ||
| Plant forestry meta‐analysis | Language + Thinning intensity | 4.23 | .049 | ||
| Language | Thinning intensity | 12.06 | .0005 | ||
Statistically significant results (in bold) indicate that effect sizes differ between English‐ and Japanese‐language studies even after controlling for the relevant fixed or random factor(s).