Urgent Help Needed- Interpretation of Categorical Variables

Thank you for taking the time to read this!!

I am really stuck with a regression I am doing for my dissertation. Basically it is the categorical variable that is confusing me. The categorical variable 'industry' has three levels - yes, no, or next (indicating whether there is or isnt industry, or whether there is industry in the next state) I also have two other independent variables. My dependent variable is Forest Loss.

When I run the regression in R industrynext 'disappears' (is this something to do with it being a reference var? I really dont know). industryno is significant and industryyes is not.

How would I interpret industryyes and industryno, and what/ where is industrynext??

I hope this has made sense!!

Thank you for any help.

Re: Urgent Help Needed- Interpretation of Categorical Variables

Hey larrikinlover.

What is your specific model that you supplied to R?

Re: Urgent Help Needed- Interpretation of Categorical Variables

Hello my model is -

lm(ForestCoverKm2~Industry+CitySize+DistanceBelem+ FedRoad)

And the output is -

Residuals:

Min 1Q Median 3Q Max

-20416 -6273 -249 2585 124963

Coefficients:

Estimate Std. Error t value Pr(>|t|)

(Intercept) -8.308e+03 2.630e+03 -3.159 0.00195 **

Industryno 4.733e+03 3.471e+03 1.364 0.17489

Industryyes 5.391e+03 2.794e+03 1.930 0.05571 .

CitySize 2.027e-03 1.130e-02 0.179 0.85793

DistanceBelem 2.357e+01 4.539e+00 5.194 7.29e-07 ***

FedRoad 5.535e+03 2.607e+03 2.124 0.03550 *

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 14410 on 137 degrees of freedom

Multiple R-squared: 0.2518, Adjusted R-squared: 0.2245

F-statistic: 9.222 on 5 and 137 DF, p-value: 1.382e-07

Re: Urgent Help Needed- Interpretation of Categorical Variables

Is there anyone that can help with this?

The deadline for the paper is tomorrow- I would really appreciate any advice.

Re: Urgent Help Needed- Interpretation of Categorical Variables

You need to setup of dummy variables for categorical regression.

Set up two dummy variables where you have:

A B

0 0 - First level

1 0 - Second level

1 1 - Third Level

where A and B are dummy variables with each level being a realization of a particular category.

Then each estimated parameter corresponds to the average number of realizations for that category.

Have you taken a course or read up on Generalized Linear Models/Linear Models before?