I know there are some statistics buffs in here

Daoist · 04-23-2012, 07:06 AM

For a regression of the form y = a + bx, you want to test whether x has any significant effect on y at all, by using a t-test (or more generally, an F test). But how about y = a + bxz? How do you differentiate the effect x has from that of z?

And a related question, if anyone is familiar with "conditional CAPM" and wants to try to convince me that it is anything more than a really sophisticated attempt by academics to troll people who don't know how to properly specify their hypothesis tests, I welcome you to try to convince me.

**Latro** · 04-23-2012, 08:49 PM

To view links or images in this forum your post count must be 2 or greater. You currently have 0 posts.

In particular the "regression statistics" section.

Daoist · 04-24-2012, 08:09 AM

Right thanks. That doesn't make too much sense to me, but at least I have a term to search for.

By the way, I take back what I said about conditional CAPM - it works in its original form. Just not the way we're doing it in this project I'm working on.

04-23-2012, 07:06 AM	#1
Daoist Veteran Member [77%] MBTI: INTP Join Date: Aug 2011 Posts: 3,085	For a regression of the form y = a + bx, you want to test whether x has any significant effect on y at all, by using a t-test (or more generally, an F test). But how about y = a + bxz? How do you differentiate the effect x has from that of z? And a related question, if anyone is familiar with "conditional CAPM" and wants to try to convince me that it is anything more than a really sophisticated attempt by academics to troll people who don't know how to properly specify their hypothesis tests, I welcome you to try to convince me.
Daoist is online post a comment

04-23-2012, 08:49 PM	#2
Latro Veteran Member [85%] MBTI: INTP Join Date: Apr 2009 Posts: 3,410	To view links or images in this forum your post count must be 2 or greater. You currently have 0 posts. In particular the "regression statistics" section.
Latro is offline post a comment

04-24-2012, 08:09 AM	#3
Daoist Veteran Member [77%] MBTI: INTP Join Date: Aug 2011 Posts: 3,085	Right thanks. That doesn't make too much sense to me, but at least I have a term to search for. By the way, I take back what I said about conditional CAPM - it works in its original form. Just not the way we're doing it in this project I'm working on.
Daoist is online post a comment