Control for yearly effects

#1
Hi everybody,

of course, I tried to find a thread with an answer about my topic. Unfortunately without success.

I ran a multi regression analysis with 8 control variables and one explanatory variables to explain my dependent variable. The date is compiled between the years 1990 and 2000.
Having done that, I found some sources, where it is said: "We use yearly dummy variables to control for yearly effects."

What does that exactly mean? Shall I create dummy variables for each year (11 additional dummies) and put them into my analysis? Or what is it all about "controlling for yearly effects?"

Perhaps my question is on a pretty low level. ;-/

Thanks for any support!

Regards
 

jpkelley

TS Contributor
#2
Good question. Most statistics programs automatically create dummy variables for you during the analysis. This is only for categorical variables (e.g. male versus female, etc.). So, if you want to treat year as a categorical factor, then simply specify "year" as such in whatever program you're using. But you might want to consider how your dependent variable changes with time, in which case you would specify year as a continuous variable.
 
#3
Thanks for your reply!

In my dataset it is a cross-sectional regression analysis. So in my case, I compiled data between these previously mentioned years and want to make an aggregate conclusion about the interference of the dependent variable and the dependent variable. Thus, I don't want to see a development of the dependent variable.

Do I need this control for yearly effects?

p.s: I'm using SPSS as well as R.