Stata-style result order

Stata-style result order

Build 2026-10-24.23 keeps the existing colors, typography, cards, controls and calculations. Reordering occurs in the document structure as well as the rendered reading order.

  • OLS: Source / SS / df / MS table for classical fits; observations used, F and its df, probability, R-squared, adjusted R-squared and Root MSE; coefficient table with the constant last. Robust/design fits omit the classical source table.
  • Logit, ordered logit and Poisson: observations, classical likelihood-ratio test where applicable, pseudo R-squared and log likelihood; estimates then notes and diagnostics. Robust/design summaries do not substitute an LR test for a Wald/design test.
  • Coefficient columns: estimate, SE, t/z, p, confidence interval, then supplementary beta/odds-ratio fields. Original engine parameter order remains intact for prediction and covariance calculations.
  • Descriptives: observations, mean, SD, minimum and maximum before additional statistics.
  • T-tests: sample/group rows with Obs, Mean, Std. err., Std. dev. and CI; difference row for paired/group designs; test statistics; chart and interpretation. Paired descriptives use only complete pairs. Welch estimation remains Welch, not Stata’s default equal-variance test.
  • Correlation, ANOVA/group comparisons and crosstabs: statistical tables before their charts.
  • Time series: sample/model information and coefficient estimates before exploratory diagnostics and charts.
  • SEM: sample/method, estimates, goodness-of-fit diagnostics, path diagram and interpretation.
  • Networks: network counts, model estimates and centrality tables before the network plot.
  • Text: counts and frequency table before the word cloud. Methods without a direct Stata equivalent retain their own result labels.
  • Quick interpretation cards follow primary results. Copy/export and publication-table ordering preserve the same coefficient sequence; full-precision t-test metadata remains available in copied output.

Validation: node calculator/tests/stata-output.test.mjs PYTHONDONTWRITEBYTECODE=1 python3 calculator/tests/stata-template.test.py

Primary references: Stata manuals for regress, logit, ttest, summarize, correlate and SEM.