This dialog can be used to specify the initial values for the covariance matrix across the random effect terms for a random coefficient regression model. If linear random coefficient terms are being used then you will need to supply the values for the variances of the intercept and linear terms, and the covariance between the two. Similarly, if linear plus quadratic terms are being fitted you must also supply the values for the variance of the quadratic term, plus the covariances for the quadratic term with the intercept, and the quadratic term with the linear term.
| These values should be supplied when fitting linear or linear plus quadratic terms | |
| Variance intercept | Value for the variance of the intercept |
| Variance linear | Value for the variance of the linear term |
| Covariance intercept/linear | Value for the covariance between intercept and linear term |
| These values should be supplied when fitting linear plus quadratic terms | |
| Variance quadratic | Value for the variance of the quadratic term |
| Covariance intercept/quadratic | Value for the covariance between intercept and quadratic term |
| Covariance linear/quadratic | Value for the covariance between linear and quadratic terms |
See also
- Random Coefficient Regression menu
- Options for specifying output options