Reference collection
Assumptions & data shape
Interpret variance, normality, multicollinearity, and other diagnostics that affect whether results can be trusted.
-
Interpretation guide
VIF and Durbin-Watson Interpretation: Competing Diagnostic Cutoffs
Interpret VIF and Durbin-Watson diagnostics with attributed competing rules, then inspect design, residuals, and coefficient stability.
-
Interpretation guide
Homoscedasticity vs Heteroscedasticity: Reading the Evidence
Interpret constant and changing residual variance using residual plots, Levene’s test, and the Breusch–Pagan test without overreading one p-value.
-
Interpretation guide
Skewness and Kurtosis: What Counts as Acceptable?
Interpret skewness, raw and excess kurtosis, sample-size effects, and distribution plots without treating floating cutoffs as universal laws.
-
Interpretation guide
Normality Test Interpretation: Shapiro-Wilk, K-S, and Anderson-Darling
Interpret Shapiro-Wilk, Kolmogorov-Smirnov, and Anderson-Darling results without treating a large p-value as proof of normality.