Hello,
I am interested in estimating a time series regression model using aggregate monthly data. For example, y = log monthly wages as a function of a vector of covariates also measured at the monthly level, with t = months.
For continuous covariates, say mean years of education, I could include this directly as a continuous variable. So, log mean wages = mean years of education.
But how would I include categorical variables?
For a binary variable, such a gender, can I include percent women as a covariate? So, log mean wages = mean years of education + percent women?
For a nominal variable with multiple categories such as race white vs black vs other, would I include percent white + percent black and omit percent other as the reference category? So, log mean wages = mean years of education + percent women + percent white + percent black?
I see dummy variables discussed briefly here (https://otexts.com/fpp3/useful-predictors.html) but not in terms of regression models with covariates.
Thank you
I am interested in estimating a time series regression model using aggregate monthly data. For example, y = log monthly wages as a function of a vector of covariates also measured at the monthly level, with t = months.
For continuous covariates, say mean years of education, I could include this directly as a continuous variable. So, log mean wages = mean years of education.
But how would I include categorical variables?
For a binary variable, such a gender, can I include percent women as a covariate? So, log mean wages = mean years of education + percent women?
For a nominal variable with multiple categories such as race white vs black vs other, would I include percent white + percent black and omit percent other as the reference category? So, log mean wages = mean years of education + percent women + percent white + percent black?
I see dummy variables discussed briefly here (https://otexts.com/fpp3/useful-predictors.html) but not in terms of regression models with covariates.
Thank you
