# Cant do a linear regression on a subset of my data: Error

#### Latriuz

##### New Member
I installed the R package AER, from which you get the data PSID1982. Then I define this model:

data("PSID1982")
set.seed(15606)
selectedobs = sample.int(nrow(PSID1982),size = 400, replace = FALSE)
attach(PSID1982);
experiencesq = experience^2
dgender = 1*(gender=="male")
dmarried = 1*(married=="yes")
dunion = 1*(union=="yes")
dindustry = 1*(industry=="yes")

outreg = lm(log(wage) ~ experience + experiencesq + education + dgender + dmarried + dunion + dindustry)
summary(outreg);

And now I need to find out if the impact of the amount of work experience on the log(wage) is different for women compared to men, and what do I expect it to be intuitively?

I want to divide the set of data into 2 subsets being men and women and perform a linear regression each dataset, so i do this. but then i get this error message and i dont know how to fix this....

> outreg = lm(wage~ experience+experiencesq+education+dmarried+dunion+dindustry, subset(PSID1982,gender=="male"))
Error in model.frame.default(formula = wage ~ experience + experiencesq + :
variable lengths differ (found for 'experiencesq')