[Alex Mueller]

Carlo,
thanks again for our reply!

: other things being equal, being a medium size company significantly explains variation in the dependent variable vs large size Company;

Is there a diffence between "explain Variation in dependent variable" and "has a significant effect"?

3) Since the coeeficient is negative this implies if i am male (sex=1) a Switch to large corporation (medium size=1 Switches to medium_size=0) lowers my Output?
What would be the Interpretation for woman (sex=0) in this case?

I will try your -predict-advise and use the delimiters, thanks for the hint

Best, A.

[Carlo Lazzaro]

Alex:
- no substantive difference between the two statements.
As far as your last question is concerned, I think that the best approach is calculating predicted by hand for the (2x2)=4 instances (please check what follows with -predicted- results):

Male and large size enterprise:
 
predicted=.3954576 + [-.0034122+(-.2141357*0)]=0.3920454     
 
Male and medium size enterprise:
 
predicted=.3954576 + [-.0034122+(-.2141357*1]=0.1779097     
 
Female and large size enterprise:
 
predicted=.3954576 + [(-.0034122*0)+(-.2141357*0)]=.3954576
 
Female and medium size enterprise:
 
predicted=.3954576 + [(-.0034122*0)+(-.2141357*0)]=.3954576

[Alex Mueller]

Hi Carlo,

thanks again,

to make the interpretation easier i plotted the margins: (please see file attached)

For me it does imply:

1) Sex alone doesnt have a significant effect
2) Large or Medium company makes a difference

3) Interaction: If I am in a medium company (medium=1), being female (sex = 0) has a positive effect, weheras being male (sex = 1) has a negative effect on my output.

Am I right in that interpretation?

[Alex Mueller]

Hi together,
can anybody confirm my Interpretation from above?

Cheers, A.

[Alex Mueller]

Noone has any idea?