Why analysts choose Statistico
Instead of isolated statistical outputs, Statistico keeps every stage of Logistic Regression connected — without leaving the analytical workflow.
From fitted model to statistical reasoning — this is Interactive Statistical Computing, not a sequence of static outputs.
What the module provides
One model. Six connected views.
Click a stage to inspect it — each view stays inside the same fitted model, so there's no re-running the analysis to move between them.
logistic-regression-results.webpResults
Model Info and Model Fit cards report −2LL, deviance, AIC, BIC, and pseudo-R², alongside the full Logistic Coefficients table — log-odds, odds ratios, and significance in one surface.
logistic-regression-predictions.webpPredictions
The Confusion Matrix and Prediction Table connect predicted classifications back to the fitted coefficients, so performance can be evaluated case by case, not just as a single summary statistic.
logistic-regression-diagnostics.webpDiagnostics
Deviance residuals, Cook's distance, and DFBETA surface influential observations, so model stability can be assessed alongside overall fit.
logistic-regression-roc.webpROC / AUC
The ROC curve and AUC pair with an adjustable classification threshold, so sensitivity, specificity, precision, recall, and F1 can be reviewed as trade-offs rather than a fixed cutoff.
logistic-regression-descriptives.webpDescriptives
Descriptive and bivariate checks summarize the outcome against each predictor and flag sparse cells, catching separation issues before they distort the fitted model.
logistic-regression-ai.webpAI Assessment
AI-assisted assessment summarizes fit, coefficients, and diagnostics in plain language and points to checks worth reviewing next — see the caution below.
Classification is part of the model workflow
The ROC curve and AUC sit next to an adjustable classification threshold — moving it updates the confusion matrix and metrics together, making the sensitivity/specificity trade-off visible rather than fixed at 0.5.
logistic-regression-roc.webpAssess fit, stability, and influential observations
Fit is reported through several complementary measures, and diagnostic views flag observations worth a second look.
Move from fitted model to prediction scenarios
Predictor-specific controls build a scenario and return the predicted probability and classification at the current threshold — drawn directly from the fitted coefficients, so scenarios can be compared without leaving the workflow.
Interpretation within the analytical workflow
Statistico's AI-assisted assessment summarizes findings, highlights relevant diagnostics, and suggests next steps — drawing on the same fit statistics and output already in the workspace.
Built for analysts who already know the method
Excel remains the working data environment.
Statistical outputs remain visible and inspectable.
Interaction supports analysis rather than hiding it.