Table 2. Results of the best fit GLMM during the proficiency phase.
Fixed effects in bold had a significant influence on whether or not an individual approached. In both dyadic and triadic sessions, individuals were more likely to approach others close in rank to themselves, unless the potential partner was kin.
| Variable | β | SE | Z | p |
|---|---|---|---|---|
| Dyadic proficiency | ||||
| Fixed effects | ||||
| Intercept | −1.98 | 0.46 | −4.30 | <0.001 |
| Kin | −1.38 | 0.67 | −2.07 | 0.04 |
| Affiliation | 0.12 | 0.10 | 1.22 | 0.22 |
| Rank distance | −0.32 | 0.08 | −3.86 | <0.001 |
| Kin ∗ rank distance | 0.52 | 0.13 | 3.80 | <0.001 |
| Affiliation ∗ rank distance | −0.03 | 0.03 | −1.12 | 0.26 |
| Random effects | ||||
| Individual present | Variance | 0.26 | ||
| Individual approaching | Variance | 0.23 | ||
| Session | Variance | 0.00 | ||
| Triadic proficiency | ||||
| Fixed Effects | ||||
| Intercept | −1.79 | 0.35 | −5.11 | <0.001 |
| Kin | −0.47 | 0.45 | −1.05 | 0.30 |
| Affiliation | −0.02 | 0.05 | −.26 | 0.80 |
| Rank distance | −0.02 | 0.04 | −0.58 | 0.56 |
| Kin ∗ rank distance | 0.22 | 0.08 | 2.67 | 0.007 |
| Affiliation ∗ rank distance | 0.02 | 0.01 | 1.26 | 0.21 |
| Random effects | ||||
| Individual present | Variance | 0.21 | ||
| Individual approaching | Variance | 0.19 | ||
| Session | Variance | 0.08 |