$p$-value based variable selection, with data that are pure noise | $p$-values, vs number of variables available for selection | Reference

Calibration of Regression Diagnostics | A 'simple' straight line regression example | The function plotSimScat() | Diagnostic Plots for Simulated Data -- plotSimDiags() | Residuals versus fitted values | A check for normality: | Is the variance constant? | Issues of leverage | All 4 default diagnostic plots in the same call | What further checks and tests should be applied? | References