I am doing a linear regression analysis in R with logarithmic dependent variable. One of the control variables is categorical and describes an industry. There are 6 industries and thereby R automatically creates a separate coefficient for each Industry. However, I just want to see if that variable is significant or not and am not interested in separate coefficient for each factor.

Is there a way to program it in a way that the summary of the model just says if the influence of industry was significant for at least one factor or not?


closed as off-topic by mkt, Peter Flom Mar 25 at 12:06

This question appears to be off-topic. The users who voted to close gave this specific reason:

  • "This question appears to be off-topic because EITHER it is not about statistics, machine learning, data analysis, data mining, or data visualization, OR it focuses on programming, debugging, or performing routine operations within a statistical computing platform. If the latter, you could try the support links we maintain." – mkt, Peter Flom
If this question can be reworded to fit the rules in the help center, please edit the question.

  • $\begingroup$ R should be creating 6-1=5 coefficients unless you are fitting a model without an intercept.. $\endgroup$ – StatsStudent Mar 23 at 18:38
  • 1
    $\begingroup$ If I understand, you want to use the Anova function in the car package. $\endgroup$ – Sal Mangiafico Mar 24 at 3:58
  • $\begingroup$ Thats Right, R creates 6-1 coefficients but I just want 1 coefficient to tell me whether the influence of any category is significant. I don't want to know which one specfically $\endgroup$ – david khubua Mar 24 at 19:05

It only makes sense to calculate a single coefficient for an independent variable if the following conditions are met:

  1. The values of the variable are ordered; i.e. each increment/decrease of one value indicates an increase/decrease in the represented phenomenon.

  2. The values of the variable are equidistant. i.e any increment of one level on the variable indicates a similar increase in the represented phenomenon.

Continuous variables often meet these criteria, but categorical variables only do in certain cases: if the categories are ordinal and equidistant (e.g., number of production facilities, income strata etc.).

I suppose your variable 'Industry' is nominal instead of ordinal, so calculating a single regression coefficient instead of separate coefficients for each of the 5 dummy-variables would not make sense.Instead, you could consider an omnibus test to test for overall differences across categories (e.g. ANOVA).


Not the answer you're looking for? Browse other questions tagged or ask your own question.