Mexican Women's Labor-Force Participation

Data from the National Survey of Household Income and Expenditures for 1977, Secretaria de Programacion y Presupuesto, Mexico.

data("MexicanLabor")

Format

A data frame containing 16 observations on 6 variables.

total

integer. Number of women older than 12 years.

laborforce

integer. Number of women in labor force.

locality

factor with levels "rural"/"urban".

age

factor with levels "<= 24" and "> 24" (in years).

income

factor with levels "low"/"high" (household income less or more than $2626.8).

schooling

factor with levels "primary" (primary school or less) and "further" (more than primary school).

Details

The data were first analyzed by Guerrero and Johnson (1982) as an example of a highly asymmetric data set, i.e., the observed proportions are rather low.

Source

Guerrero V, Johnson R (1982). “Use of the Box-Cox Transformation with Binary Response Models.” Biometrika, 69, 309–314.

Examples

## data
data("MexicanLabor", package = "glmx")

## visualizations
plot(I(laborforce/total) ~ interaction(income, age), data = MexicanLabor)

plot(I(laborforce/total) ~ interaction(schooling, locality), data = MexicanLabor)


## simple logit model
m <- glm(cbind(laborforce, total - laborforce) ~ ., data = MexicanLabor, family = binomial)
m
#> 
#> Call:  glm(formula = cbind(laborforce, total - laborforce) ~ ., family = binomial, 
#>     data = MexicanLabor)
#> 
#> Coefficients:
#>      (Intercept)     localityurban           age> 24        incomehigh  
#>         -2.58575           0.38931           0.29212          -0.01521  
#> schoolingfurther  
#>          0.54814  
#> 
#> Degrees of Freedom: 15 Total (i.e. Null);  11 Residual
#> Null Deviance:	    73.89 
#> Residual Deviance: 15.96 	AIC: 95.61