Fit Binary Logistic Model - Options
main topic     see also      

Stat > Regression > Binary Logistic Regression > Fit Binary Logistic Model > Options

Provides options that allow you to choose a link function, specify weights, set the confidence level, choose the type of confidence interval, choose residuals, choose the type of deviances to use for the deviance tests, and set the number of groups for the Hosmer and Lemeshow test.

Dialog box items

Link Functions Minitab provides three link functions, allowing you to fit a broad class of binary response models. The link function is the inverse of a distribution function. You want to choose a link function that results in a good fit to your data. Goodness-of-fit statistics can be used to compare the fits using different link functions. Certain link functions may be used for historical reasons or because they have a special meaning in a discipline. An advantage of the logit link function is that it provides an estimate of the odds ratios. For a comparison of link functions, see [25].

Logit: Choose to use the logit link function. The logit link function is the canonical link function for binary logistic regression.

Normit/Probit: Choose to use the normit link function.

Gompit/Complementary log-log: Choose to use the gompit link function (also called the complementary log-log function). The gompit function is the inverse of the Gompertz distribution function.

Weights: Enter a numeric column of weights to perform weighted regression. Weights must be greater than or equal to zero. The length of the weights column must match the length of the response column.

Confidence level for all intervals: Enter the confidence level. The default is 95.

Type of confidence interval: Choose the type of confidence interval: two-sided (default), lower bound, or upper bound.

Residuals for diagnostics: Choose between Pearson and Deviance residuals. Either version helps to find patterns in the residual plots and outliers. When the model uses the logit link function, the distribution of the deviance residuals is more like the distribution of residuals from least squares models.

Deviances for Tests: Select a deviance for calculating chi-square values and p-values. Typically, you use the adjusted deviance. Use sequential deviance  to determine the significance of terms by the order that they enter the model.

Adjusted (Type III): Measures the reduction in the deviance for each term relative to a model that contains all of the remaining terms.  

Sequential (Type I):  Measures the reduction in the deviance when a term is added to a model that contains only the terms before it.

Number of Groups for Hosmer-Lemeshow Test : Enter the number of groups for the Hosmer-Lemeshow test. The default is 10. See [24] for details.