Logistic regression software with odds ratios and diagnostics Binary logistic and probit regression — parameter estimates, odds ratios with confidence intervals, AIC/BIC, Type I and Type III term-level tests, leverage plots, and influence diagnostics. Build, examine, and refine the model in one iterative workflow.
Try free for 15 days — no credit card required
Trusted by 75,000 researchers, analysts, and scientists at leading universities, hospitals, and companies worldwide — for over 30 years.
Complete logistic regression, not just coefficients
Binary logistic and probit regression with continuous variables, categorical factors, and interaction terms. Odds ratios with confidence intervals, Type I and Type III tests, leverage plots, AIC, BIC, and the model equation — in a single integrated workflow.
Iterate until the model is right
Add or remove predictors, test interactions, inspect influence diagnostics, and re-run immediately. When Cook's D flags an influential observation or a predictor adds nothing, adjust the model and see the effect. The process stays iterative, not one-shot.
NIST-validated accuracy
All calculations run in Analyse-it's own validated engine — not Excel's built-in functions. Validated against published reference datasets with IEEE 754 double-precision arithmetic. The numbers you report can be defended under peer review or audit.
Logistic regression is not available in Excel's built-in Analysis ToolPak — it stops at linear regression, and there is no workaround that produces the output practitioners actually need: odds ratios, confidence intervals, model fit statistics, term-level hypothesis tests, and diagnostics. For most researchers and analysts whose data lives in Excel, this has meant either learning a programming language or buying a separate application with its own interface, licence, and file format — for an analysis that should sit naturally alongside the data.
Analyse-it treats logistic regression the same way it treats linear regression: as a process of building, examining, and refining a model. Start with your data, choose your predictors — continuous variables, categorical factors, polynomial terms, interactions — and get a complete analysis. Parameter estimates with standard errors, odds ratios with confidence intervals, Wald and likelihood ratio tests for the model and for each term individually, AIC, BIC, log-likelihood, and deviance. Leverage plots to see each predictor's contribution after accounting for all others. Residual diagnostics and the Outlier and Influence plot using Cook's D and Studentised Residuals to identify observations that have a disproportionate effect on the model.
When a diagnostic suggests a problem — an influential observation, a predictor that adds nothing, an interaction worth testing — adjust the model and re-run immediately. The results update in the same workbook. Categorical variables are coded automatically as dummy variables. The model equation is produced for direct use in documentation, reports, or supplementary materials. Predict new observations directly from the fitted model, or save fitted values, residuals, leverage, and Cook's D back to the dataset for further analysis.
75,000 researchers, analysts, and scientists at leading universities, hospitals, and companies worldwide have relied on Analyse-it for over 30 years.
What's included
-
Fit binary logistic and probit models
Logistic regression using the logit link function, or probit regression using the normal CDF link. Simple and multiple predictor models with continuous variables, categorical factors, polynomial terms, and interactions. Categorical variables coded automatically — no manual recoding.
-
Report odds ratios and model fit
Odds ratio estimates with confidence intervals and significance tests for each predictor. AIC, BIC, log-likelihood, and deviance for comparing competing models. Parameter estimates with standard errors and Wald statistics. The model equation for documentation.
-
Assess each predictor's contribution independently
Sequential (Type I) and partial (Type III) hypothesis tests for every term. Leverage plots to visualise the effect of each predictor after accounting for all others. Correlation and covariance of estimates matrices for assessing multicollinearity.
-
Identify influential observations before they mislead
Residual plots, distribution plots, and sequence plots. Outlier and Influence plot using Cook's D and Studentised Residuals — know whether your conclusions depend on a handful of unusual cases.
-
Predict new observations from the fitted model
Save fitted values, residuals, standardised residuals, studentised residuals, leverage, and Cook's D back to the dataset. Predict new observations directly from the model without re-entering data.
-
Part of the Standard edition
Logistic regression is part of the Standard edition, which also covers simple and multiple regression, ANOVA and ANCOVA, PCA and factor analysis, descriptive analysis, hypothesis testing, and categorical data analysis. All the professional statistics you need — at a fraction of the cost of dedicated packages.
Trusted by 75,000 researchers, analysts, and scientists at leading universities, hospitals, and companies worldwide — for over 30 years.
Try free for 15 days — no credit card required
