dummy command

[Klaudia Lennerling]

Hello!

I've been using stata and doing regressions. One of my control variables is a categorical variable divided into different sectors so when I did my regression I did the i.sector command to divide it into dummies. However, the table only shows the 2 last sectors and does not include the first sector. Did I do it the right way or do I have to do something else? And if I did it the right way how do I interpret the result when the first category is not included?

Best regards, Klaudia

[Carlo Lazzaro]

Klaudia:
by default -fvvarlist- notation shelters your regression from the so called "dummy trap" (

https://en.wikipedia.org/wiki/Dummy_variable_(statistics)

) by omitting one level of the categorical (which sounds better than dummy) variables (which acts as the reference category).
As far as the interperetation of your resulst is concerned, even guess-work capabilities have their own limits : as per FAQ, please share (via CODE delimiters) what you tyoed and what Stata gave you back. Thanks.

[Fei Wang]

You did it right. The first category is omitted due to perfect collinearity. The coefficients of the remaining two categories measure the differences of Y between category 2 (or 3) and category 1, conditional on other covariates. In other words, the omitted category is a base group with which other groups are compared.

Crossed with Carlo