CODEWITHSUNDEEP
testingregressionstatacategorical-dataeconomics
user3466971
user3466971

Reputation: 5

Testing if two independent variables have the same effect

I have variables age (beginning from 65) and female (1 if female and 0 if male) in my dataset.

How can I test if being male has the same effect as aging 10 years on the total health expenditure?

I have written something like this, which is probably incorrect:

regress totexp age female
test i0.female = age / 10

A reproducible example using Stata's auto toy dataset is the following:

sysuse auto
regress price i.foreign mpg
test i0foreign = mpg / 10

Upvotes: 0

Views: 506

Answers (1)

dimitriy
dimitriy

Reputation: 9470

You can try something along the following lines:

regress price i.foreign c.mpg

lincom _b[1.foreign] - (_b[mpg] / 25)
test   _b[1.foreign] = (_b[mpg] / 25)

Note that 25 is an arbitrary number to illustrate the idea. You can also divide age by 10 before fitting the regression to put it into decades. In addition, you probably want to multiply rather than divide in your original example if age is measured in years.

Upvotes: 1

Related Questions