Logistic Regression

This web page performs logistic regression, in which a dichotomous outcome is predicted by one or more variables. The program generates the coefficients of a prediction formula (and standard errors of estimate and significance levels), and odds ratios (with confidence intervals).
It extends John Pezzullo's http://statpages.info/logistic.html enabling you to calculate the predictive probability for specific covariate pattern(s), and for one dimensional models, ie. models with ONLY ONE predictor variable, you can generate a plot. You may optionally use summary data in either of 2 ways

a variable with weights or
two variables, one with the numbers of success and one with the number of failures

These options is demonstrated in two of the demo dataset.

DATA

N-row × M-col matrix
First row must contain column names
Use short informative names without special characters
Delimiter comma, space or tab
Use NA for missing cases
Write, copy/paste or load CSV file

MODEL

Specify model (watch examples)
List (comma separated) categorial variables
Alpha-Numeric variable is treated as categorial
For weigthed summary data specify the weight-variable

Options:

Model: Dichotomous outcome variable ~ Predictor variable(s)

Type Operator Example

Addive terms + varA + varB

Interaction terms : varA : varB

Both * varA + varB + varA * varB

Keep as is I() I(varA^2), I(varA**varA) and I(varA*varA)
all expressions squares varA

ℹ

Model

Categorial variable(s)

Weight

Conf.level

Plot

Only 1-dimensional models will plot

Home