lm2.md
In statuser: Statistical Tools Designed for End Users

lm2 print output format is stable

Code
  print(result)
Output
  Call: lm2(formula = mpg ~ wt + hp, data = mtcars, notes = FALSE)

              estimate  SE.robust SE.classical t.value p.value std.estimate
  intercept    37.23**    2.230      1.599      16.70   <.0001        --   
  wt          -3.878**    0.769      0.633     -5.046   <.0001    -0.630   
  hp          -0.032**    .0094      .0090     -3.385    .0021    -0.361   
                mean missing red.flag
  intercept      --    --        !   
  wt          3.217    0         --  
  hp          146.7    0         --

  N = 32  | missing = 0  | df = 29  | R² = 0.827  | SE type: HC3

lm2 print output with notes is stable

Code
  print(result)
Output
  Call: lm2(formula = mpg ~ wt, data = mtcars, notes = TRUE)

              estimate  SE.robust SE.classical t.value p.value std.estimate
  intercept    37.29**    2.427      1.878      15.36   <.0001        --   
  wt          -5.344**    0.738      0.559     -7.241   <.0001    -0.868   
                mean missing red.flag
  intercept      --    --        !   
  wt          3.217    0         !

  N = 32  | missing = 0  | df = 30  | R² = 0.753  | SE type: HC3

  Notes:
    - ** p<.01 (based on SE.robust)
    - std.estimate is the standardized coefficient: beta = b * sd(x) / sd(y)
    - mean: for numeric variables, mean of x; for factors, % of observations
    - missing: number of observations excluded due to missing values
    - red.flag:
       !: robust & classical SE differ by more than 25%
       - Suggestion 1: to evaluate possible extreme skew or outliers:
         plot_density() or plot_freq() for mpg and for wt
       - Suggestion 2: to evaluate possible nonlinearity, do plot_gam(mpg ~ wt)
    - To avoid these notes, lm2(..., notes=FALSE)

lm2 print output with interaction is stable

Code
  print(result)
Output
  Call: lm2(formula = mpg ~ wt * hp, data = mtcars, notes = FALSE)

              estimate  SE.robust SE.classical t.value p.value std.estimate
  intercept    49.81**    5.113      3.605      9.741   <.0001        --   
  wt          -8.217**    1.655      1.270     -4.965   <.0001    -1.334   
  hp          -0.120**    0.032      0.025     -3.732    .0009    -1.366   
  wt:hp        0.028**    .0096      .0074      2.888    .0074     0.310   
                mean missing r(x,z)   red.flag
  intercept      --    --          --     !   
  wt          3.217    0           --     !   
  hp          146.7    0           --     !   
  wt:hp       514.7    --      0.66**     ! X

  N = 32  | missing = 0  | df = 28  | R² = 0.885  | SE type: HC3