Print estimated coefficients and their standard errors in a table for several regression models.

This function extracts estimates of regression parameters and their standard errors from one or more models and prints them in a table.

Usage

compareCoefs(..., se = TRUE, zvals = FALSE, pvals = FALSE, vcov.,
    print = TRUE, digits = 3)

Arguments

...: One or more regression-model objects. These may be of class lm, glm, nlm, or any other regression method for which the functions coef and vcov return appropriate values, or if the object inherits from the mer class created by the lme4 package or lme in the nlme package.
se: If TRUE, the default, show standard errors as well as estimates.
zvals: If TRUE (the default is FALSE), print Wald statistics, the ratio of each coefficient to its standard error.
pvals: If TRUE (the default is FALSE), print two-sided p-values from the standard normal distribution corresponding to the Wald statistics.
vcov.: an optional argument, specifying a function to be applied to all of the models, returning a coefficient covariance matrix for each, or a list with one element for each model, with each element either containing a function to be applied to the corresponding model or a coefficient covariance matrix for that model. If omitted, vcov is applied to each model. This argument can also be a list of estimated covariance matrices of the coefficient estimates.
print: If TRUE, the default, the results are printed in a nice format using printCoefmat. If FALSE, the results are returned as a matrix
digits: Passed to the printCoefmat function for printing the result.

Value

This function is mainly used for its side-effect of printing the result. It also invisibly returns a matrix of estimates, standard errors, Wald statistics, and p-values.

References

Fox, J. and Weisberg, S. (2019) An R Companion to Applied Regression, Third Edition, Sage.

Author

John Fox jfox@mcmaster.ca

Examples

mod1 <- lm(prestige ~ income + education, data=Duncan)
mod2 <- update(mod1, subset=-c(6,16))
mod3 <- update(mod1, . ~ . + type)
mod4 <- update(mod1, . ~ . + I(income + education)) # aliased coef.
compareCoefs(mod1)
#> Calls:
#> lm(formula = prestige ~ income + education, data = Duncan)
#> 
#>             Model 1
#> (Intercept)   -6.06
#> SE             4.27
#>                    
#> income        0.599
#> SE            0.120
#>                    
#> education    0.5458
#> SE           0.0983
#>                    
compareCoefs(mod1, mod2, mod4)
#> Calls:
#> 1: lm(formula = prestige ~ income + education, data = Duncan)
#> 2: lm(formula = prestige ~ income + education, data = Duncan, subset = -c(6,
#>    16))
#> 3: lm(formula = prestige ~ income + education + I(income + education), data 
#>   = Duncan)
#> 
#>                       Model 1 Model 2 Model 3
#> (Intercept)             -6.06   -6.41   -6.06
#> SE                       4.27    3.65    4.27
#>                                              
#> income                  0.599   0.867   0.599
#> SE                      0.120   0.122   0.120
#>                                              
#> education              0.5458  0.3322  0.5458
#> SE                     0.0983  0.0987  0.0983
#>                                              
#> I(income + education)                 aliased
#> SE                                           
#>                                              
compareCoefs(mod1, mod2, mod3, zvals=TRUE, pvals=TRUE)
#> Calls:
#> 1: lm(formula = prestige ~ income + education, data = Duncan)
#> 2: lm(formula = prestige ~ income + education, data = Duncan, subset = -c(6,
#>    16))
#> 3: lm(formula = prestige ~ income + education + type, data = Duncan)
#> 
#>             Model 1 Model 2 Model 3
#> (Intercept)  -6.065  -6.409  -0.185
#> SE            4.272   3.653   3.714
#> z             -1.42   -1.75   -0.05
#> Pr(>|z|)    0.15571 0.07932 0.96026
#>                                    
#> income       0.5987  0.8674  0.5975
#> SE           0.1197  0.1220  0.0894
#> z              5.00    7.11    6.69
#> Pr(>|z|)    5.6e-07 1.1e-12 2.3e-11
#>                                    
#> education    0.5458  0.3322  0.3453
#> SE           0.0983  0.0987  0.1136
#> z              5.56    3.36    3.04
#> Pr(>|z|)    2.8e-08 0.00077 0.00237
#>                                    
#> typeprof                      16.66
#> SE                             6.99
#> z                              2.38
#> Pr(>|z|)                    0.01722
#>                                    
#> typewc                       -14.66
#> SE                             6.11
#> z                             -2.40
#> Pr(>|z|)                    0.01639
#>                                    
compareCoefs(mod1, mod2, se=FALSE)
#> Calls:
#> 1: lm(formula = prestige ~ income + education, data = Duncan)
#> 2: lm(formula = prestige ~ income + education, data = Duncan, subset = -c(6,
#>    16))
#> 
#>             Model 1 Model 2
#> (Intercept)   -6.06   -6.41
#> income        0.599   0.867
#> education     0.546   0.332
compareCoefs(mod1, mod1, vcov.=list(vcov, hccm))
#> Standard errors computed by list(vcov, hccm)Calls:
#> 1: lm(formula = prestige ~ income + education, data = Duncan)
#> 2: lm(formula = prestige ~ income + education, data = Duncan)
#> 
#>             Model 1 Model 2
#> (Intercept)   -6.06   -6.06
#> SE             4.27    3.15
#>                            
#> income        0.599   0.599
#> SE            0.120   0.183
#>                            
#> education    0.5458  0.5458
#> SE           0.0983  0.1495
#>