Page 1 of 1

Probit regression?

Posted: Sun Apr 12, 2015 2:06 am
by harshil59
Hi, I'm currently trying to run a regression which determines what affects penalty kick decisions. I've only just found out that because I'm modelling for penalty or no penalty it is bes to use a probit estimation instead of least squares. My issue is that if I do for example run such a regression, how do you interrupt each coefficient. E.g:

http://puu.sh/h9V20/081b1733f8.png​

I know that you can analyse the significance as normal, but what about the coefficient. For example take_ons variable could you say the positive value means the more take ons, the more the penalties- but you couldnt say the extent of this as simply 0.01?

How do you interupt these results?

Also can someone explain why it is better to use a probit regression here?