I t would be inappropriate to enter OCC1 directly into a regression model. Why?
ID: 1192963 • Letter: I
Question
It would be inappropriate to enter OCC1 directly into a regression model.
Why?
The data come from a survey of 1000 individuals; these data are stored in the EViews workfile named wage. Download, save, and open this workfile in EViews following the same procedures you used in previous exercises. The data come from the Current Population Survey. It is a random sample, with replacement, of 1,000 observations from a sample of males. Nine variables relevant to a model of wage determination have been saved in this workfile: AGE (you know what) LNWAGE Log of wage (It is conventional in wage studies to use the wage variable in its natural logarithmic form. This will be the dependent variable.) OCC1 an indicator variable for occupational categories (see below). UNION 1 if union member, 0 otherwise GRADE highest educational grade completed MARRIED 1 if married, 0 otherwise PARTT 1 if part-time worker, 0 otherwise POTEXP Years of potential experience HIGH 1 if individual works in a "Highly" unionized industry The categories for OCC1 are: Managers and administrators Professionals Nurses and other non-doctors Clerical Salespeople Service workers Manual workers Craft workers This data set is typical of those derived from a survey. The unit of observation is an individual. Many of the variables, such as OCC1, are categorical: the numbers do not have cardinal, or even ordinal, meaning; rather they simply represent an occupational category. OCC1 takes on the values 1 through 8, according to the table above. It would be inappropriate to enter OCC1 directly into a regression model. Why? If you have a categorical variable that can take on more than two values, such as OCC1, you can make several dummy variables out of this categorical variable. You could simply defineExplanation / Answer
Yes, it would be inappropriate to enter OCC1 directly into the regression model. The reason being that it has 8 sub-categories and makes the model unnecessary complicated. The problem arises while interpreting results of the regression.