crs (version 0.15-31)

wage1: Cross-Sectional Data on Wages

Description

Cross-section wage data consisting of a random sample taken from the U.S. Current Population Survey for the year 1976. There are 526 observations in total.

Usage

data("wage1")

Format

A data frame with 24 columns, and 526 rows.

wage

column 1, of type numeric, average hourly earnings

educ

column 2, of type numeric, years of education

exper

column 3, of type numeric, years potential experience

tenure

column 4, of type numeric, years with current employer

nonwhite

column 5, of type factor, =“Nonwhite” if nonwhite, “White” otherwise

female

column 6, of type factor, =“Female” if female, “Male” otherwise

married

column 7, of type factor, =“Married” if Married, “Nonmarried” otherwise

numdep

column 8, of type numeric, number of dependents

smsa

column 9, of type numeric, =1 if live in SMSA

northcen

column 10, of type numeric, =1 if live in north central U.S

south

column 11, of type numeric, =1 if live in southern region

west

column 12, of type numeric, =1 if live in western region

construc

column 13, of type numeric, =1 if work in construc. indus.

ndurman

column 14, of type numeric, =1 if in nondur. manuf. indus.

trcommpu

column 15, of type numeric, =1 if in trans, commun, pub ut

trade

column 16, of type numeric, =1 if in wholesale or retail

services

column 17, of type numeric, =1 if in services indus.

profserv

column 18, of type numeric, =1 if in prof. serv. indus.

profocc

column 19, of type numeric, =1 if in profess. occupation

clerocc

column 20, of type numeric, =1 if in clerical occupation

servocc

column 21, of type numeric, =1 if in service occupation

lwage

column 22, of type numeric, log(wage)

expersq

column 23, of type numeric, exper\(^2\)

tenursq

column 24, of type numeric, tenure\(^2\)

References

Wooldridge, J.M. (2000), Introductory Econometrics: A Modern Approach, South-Western College Publishing.

Examples

# NOT RUN {
data(wage1)

## Cross-validated model selection for spline degree and bandwidths Note
## - we override the default nmulti here to get a quick illustration
## (we don't advice doing this, in fact advise using more restarts in
## serious applications)

model <- crs(lwage~married+
             female+
             nonwhite+                
             educ+
             exper+
             tenure,
             basis="additive",
             complexity="degree",
             data=wage1,
             segments=c(1,1,1),
             nmulti=1)

summary(model)

## Residual plots
plot(model)
## Partial mean plots (control for non axis predictors)
plot(model,mean=TRUE)
## Partial first derivative plots (control for non axis predictors)
plot(model,deriv=1)
## Partial second derivative plots (control for non axis predictors)
plot(model,deriv=2)
## Compare with local linear kernel regression
require(np)
model <- npreg(lwage~married+
               female+
               nonwhite+                
               educ+
               exper+
               tenure,
               regtype="ll",
               bwmethod="cv.aic",
               data=wage1)

summary(model)

## Partial mean plots (control for non axis predictors)
plot(model,common.scale=FALSE)
## Partial first derivative plots (control for non axis predictors)
plot(model,gradients=TRUE,common.scale=FALSE)
detach("package:np")
# }