Excel's t stat question.

#1
Hi guys,

I'm a little confused over this right now. I performed a regression in excel and as an example, this is what i got for my first independent variable:

X Variable 1 2.772078853 0.51663737 5.365618161 4.64369E-07

where each of the values are: coefficient, standard error, t stat, and p-value respectively.

So the t stat here is 5.365 and the p-value is 4.64E-07.

My question is, does this determine if the value is significant or not (as in a hypothesis test)?
so if t actual > t critical, does that mean i can reject the null hypothesis? I'm not too sure about this stuff so any help would be appreciated.
 

CB

Super Moderator
#2
so if t actual > t critical, does that mean i can reject the null hypothesis? I'm not too sure about this stuff so any help would be appreciated.
Yes, but you have a p value right there (which is calculated using the t-statistic and df); just use that. If the p value is less than your chosen alpha value, the coefficient is significant. The p value you have would be significant by any conventional criterion.
 
#3
Okay great! thanks for the help. So if i chose alpha as 0.05, any p-value less than that shows that the co-efficient is significant if i understand correctly.

Also is there a special case for binary variables (that can only take on 0 and 1) since i have 4 of those and it seems that all their p-values are higher than 0.05, which i guess means none of them are significant.