Select the output to be generated initially in a regression analysis – the same information can also be displayed after the analysis, using the Further output dialog. You can also request that no constant term is included in the model: this will affect only the parameterization of factor effects, if there are factors in the model; but if not, it will constrain the regression to pass through the origin.
For a general regression model, you can also control the maximum order of interaction to be generated when you use model-formula operators like *. The default is to include all interactions, up to those involving nine variates or factors. (You cannot ask for more than nine.)
|Model||Details of the model that is fitted|
|F-probabilities||F probabilities for variance ratios|
|Correlations||Correlations between the parameter estimates|
|Fitted values||Table containing the values of the response variate, the fitted values, standardized residuals and leverages|
|Estimates||Estimates of the parameters in the model|
|t-probability||t probabilities for the parameter estimates|
|Confidence intervals||Confidence intervals for the parameter estimates. The confidence limit can specified as a percentage using the Confidence limit for estimates (%) field.|
|Accumulated||Analysis of variance table containing a line for each change in the fitted model|
|Wald tests||Wald and F tests for dropping terms from a regression|
Estimate constant term
Specifies whether to include a constant in the regression model. In simple linear regression this omits the intercept, in other words the fitted line is constrained to pass through the origin.
A variate of weights can be supplied to give varying influence of each unit on the fit of the model. This would usually correspond to a known pattern of variance of the observations, when the weights would be the reciprocals of the variances.
A factor can be supplied to specify an absorbing factor defining the groups for within-groups linear regression.
Factorial limit on model terms
For General Linear Regression you can control the maximum order of interaction to be generated when you use model-formula operators like ‘*’. The default is to include all interactions, up to those involving nine variates or factors (you cannot ask for more than nine).
- Linear Regression for information on general options and other models
- Linear Regression Options for choosing which results to display
- Linear Regression Further Output for additional output subsequent to analysis
- Saving Results for further analysis
- Save Individual Regression Terms dialog
- Fitted Model for graphical display of the model
- Model Checking for diagnostic plots
- Predictions from Generalized Linear Model
- Plot Table of Predictions
- Multiple Comparisons options