Table 12 Model and result.
Model | Fixed effect | Random effect | Fixed effect significance | Random effect covariance |
|---|---|---|---|---|
\(\begin{array}{l}\log (E|Cogmech_{ti}) = \beta _{00} + \beta _{10}time_{ti} + \beta _{20}sex_{ti}\\ + \beta _{30}zone_{ti} + \beta _{40}CCI_{ti} + \beta _{50}PMG_{ti} + u_{1i}time_{ti} + u_{0i} + \varepsilon _{ti}\end{array}\) | β00 ~ β50 | u2i/u3i | β00/β40 significant (P < 0.05) | No |
\(\begin{array}{l}\log (E|Certain_{ti}) = \beta _{00} + \beta _{10}time_{ti} + \beta _{20}sex_{ti}\\ + \beta _{30}zone_{ti} + \beta _{40}CCI_{ti} + \beta _{50}Cogmech_{ti}\\ + u_{2i}sex_{ti} + u_{3i}zone_{ti} + u_{0i} + \varepsilon _{ti}\end{array}\) | β00 ~ β50 | u2i/u3i | β10/β40/β50 significant (P < 0.05) | No |
\(\begin{array}{l}\log (E|Affect_{ti}) = \beta _{00} + \beta _{10}time_{ti} + \beta _{20}sex_{ti}\\ + \beta _{30}zone_{ti} + \beta _{40}CCI_{ti} + \beta _{50}PMG_{ti} + \beta _{60}Cogmech_{ti}\\ + \beta _{70}Certain_{ti} + u_{2i}sex_{ti} + u_{0i} + \varepsilon _{ti}\end{array}\) | β00 ~ β70 | u2i | β10/β40/β60/ β70 significant (P < 0.05) | Significant gender/geographic differences |
