Oh, and I'm a dean and you're all losers.
I calculate zscores based on coefficients alone, AMA

Hahahaha wow the racist trolls are so dumb that they don't get it no matter how many times you explain it to them! It's that little something called a coefficient. Which you use when you adjust the degrees of freedom of the z scores. Understand so far?
Yeah but do you know what a test statistic is?Idiot who got schooled in the other thread and has no response detected! Thanks for admitting you were wrong!

I'm not the poster you insulted in that thread. But do give more details on your method! It's entertaining.
Everything is explained well enough there that someone with basic stats knowledge and even average intelligence can understand that I'm 100% right. So if you don't get it, what does that say?

I'm not the poster you insulted in that thread. But do give more details on your method! It's entertaining.
Everything is explained well enough there that someone with basic stats knowledge and even average intelligence can understand that I'm 100% right. So if you don't get it, what does that say?
Can you give an example of a hypothesis test using a coefficient?

I'm not the poster you insulted in that thread. But do give more details on your method! It's entertaining.
Everything is explained well enough there that someone with basic stats knowledge and even average intelligence can understand that I'm 100% right. So if you don't get it, what does that say?Can you give an example of a hypothesis test using a coefficient?
Yes. Already explained in detail in the other thread: For a regression, you compare the coefficient to a t distribution multiplied by the standard error. Second thing is still a distribution, of course. Voila, coefficient as a test statistic. Understand at long last?

I'm not the poster you insulted in that thread. But do give more details on your method! It's entertaining.
Everything is explained well enough there that someone with basic stats knowledge and even average intelligence can understand that I'm 100% right. So if you don't get it, what does that say?
Can you give an example of a hypothesis test using a coefficient?Yes. Already explained in detail in the other thread: For a regression, you compare the coefficient to a t distribution multiplied by the standard error. Second thing is still a distribution, of course. Voila, coefficient as a test statistic. Understand at long last?
The coefficient isn’t a test statistic there. The tstatistic is. The fact that you’re using algebra doesn’t change that...

I'm not the poster you insulted in that thread. But do give more details on your method! It's entertaining.
Everything is explained well enough there that someone with basic stats knowledge and even average intelligence can understand that I'm 100% right. So if you don't get it, what does that say?
Can you give an example of a hypothesis test using a coefficient?
Yes. Already explained in detail in the other thread: For a regression, you compare the coefficient to a t distribution multiplied by the standard error. Second thing is still a distribution, of course. Voila, coefficient as a test statistic. Understand at long last?The coefficient isn’t a test statistic there. The tstatistic is. The fact that you’re using algebra doesn’t change that...
Nope. Completely wrong. You're taking the coefficient and comparing it to a distribution. Hence, the coefficient is being used as a test statistic. Look at any definition of test statistic and it will qualify. I think you're coming around to the fact I'm right.

I'm not the poster you insulted in that thread. But do give more details on your method! It's entertaining.
Everything is explained well enough there that someone with basic stats knowledge and even average intelligence can understand that I'm 100% right. So if you don't get it, what does that say?
Can you give an example of a hypothesis test using a coefficient?
Yes. Already explained in detail in the other thread: For a regression, you compare the coefficient to a t distribution multiplied by the standard error. Second thing is still a distribution, of course. Voila, coefficient as a test statistic. Understand at long last?
The coefficient isn’t a test statistic there. The tstatistic is. The fact that you’re using algebra doesn’t change that...Nope. Completely wrong. You're taking the coefficient and comparing it to a distribution. Hence, the coefficient is being used as a test statistic. Look at any definition of test statistic and it will qualify. I think you're coming around to the fact I'm right.
The coefficient is the tstatistic. And the fact that the two are algebraically equivalent doesn’t mean much. Even if you were to broaden the definition it would be irrelevant. You’re still failing to engage the critique as no one would ever actually think of things in that way.

I'm not the poster you insulted in that thread. But do give more details on your method! It's entertaining.
Everything is explained well enough there that someone with basic stats knowledge and even average intelligence can understand that I'm 100% right. So if you don't get it, what does that say?
Can you give an example of a hypothesis test using a coefficient?Yes. Already explained in detail in the other thread: For a regression, you compare the coefficient to a t distribution multiplied by the standard error. Second thing is still a distribution, of course. Voila, coefficient as a test statistic. Understand at long last?
Keep making a fool of yourself CG defender troll!

I'm not the poster you insulted in that thread. But do give more details on your method! It's entertaining.
Everything is explained well enough there that someone with basic stats knowledge and even average intelligence can understand that I'm 100% right. So if you don't get it, what does that say?
Can you give an example of a hypothesis test using a coefficient?Yes. Already explained in detail in the other thread: For a regression, you compare the coefficient to a t distribution multiplied by the standard error. Second thing is still a distribution, of course. Voila, coefficient as a test statistic. Understand at long last?
So it's acceptable for you to comment on the intelligence of people who in your opinion don't understand hypothesis testing. But if someone pointed out it is *you* who don't understand, and made similar inferences on your intelligence, that would be "racist"?

I'm not the poster you insulted in that thread. But do give more details on your method! It's entertaining.
Everything is explained well enough there that someone with basic stats knowledge and even average intelligence can understand that I'm 100% right. So if you don't get it, what does that say?
Can you give an example of a hypothesis test using a coefficient?
Yes. Already explained in detail in the other thread: For a regression, you compare the coefficient to a t distribution multiplied by the standard error. Second thing is still a distribution, of course. Voila, coefficient as a test statistic. Understand at long last?
The coefficient isn’t a test statistic there. The tstatistic is. The fact that you’re using algebra doesn’t change that...Nope. Completely wrong. You're taking the coefficient and comparing it to a distribution. Hence, the coefficient is being used as a test statistic. Look at any definition of test statistic and it will qualify. I think you're coming around to the fact I'm right.
Which distribution are you comparing the coefficient to?

I'm not the poster you insulted in that thread. But do give more details on your method! It's entertaining.
Everything is explained well enough there that someone with basic stats knowledge and even average intelligence can understand that I'm 100% right. So if you don't get it, what does that say?
Can you give an example of a hypothesis test using a coefficient?
Yes. Already explained in detail in the other thread: For a regression, you compare the coefficient to a t distribution multiplied by the standard error. Second thing is still a distribution, of course. Voila, coefficient as a test statistic. Understand at long last?So it's acceptable for you to comment on the intelligence of people who in your opinion don't understand hypothesis testing. But if someone pointed out it is *you* who don't understand, and made similar inferences on your intelligence, that would be "racist"?
The racist trolls are clearly racist because they're fixated on CG and inventing false claims and bogus criticisms. It's not like this is hidden or anything. They/you are constantly pointing to her race. The fact that the lies and nonsense are so easily exposed is also clear evidence that it's motivated by racial grievance. I've been insulting people's intelligence because they're saying dumb things and pretending not to understand extremely simple arguments to save face. Sorry not sorry.

I'm not the poster you insulted in that thread. But do give more details on your method! It's entertaining.
Everything is explained well enough there that someone with basic stats knowledge and even average intelligence can understand that I'm 100% right. So if you don't get it, what does that say?
Can you give an example of a hypothesis test using a coefficient?
Yes. Already explained in detail in the other thread: For a regression, you compare the coefficient to a t distribution multiplied by the standard error. Second thing is still a distribution, of course. Voila, coefficient as a test statistic. Understand at long last?
The coefficient isn’t a test statistic there. The tstatistic is. The fact that you’re using algebra doesn’t change that...
Nope. Completely wrong. You're taking the coefficient and comparing it to a distribution. Hence, the coefficient is being used as a test statistic. Look at any definition of test statistic and it will qualify. I think you're coming around to the fact I'm right.The coefficient is the tstatistic. And the fact that the two are algebraically equivalent doesn’t mean much. Even if you were to broaden the definition it would be irrelevant. You’re still failing to engage the critique as no one would ever actually think of things in that way.
"The coefficient is the tstatistic"? What? There's no t statistic involved. You compare the coefficient to the distribution defined by the standard error times the t distribution. Boom  coefficient as a test statistic. There's also no "broadening" of the definition. Look up how a test statistic is defined if you don't understand this point.
This. Is. Extremely. Simple. If you don't admit you're wrong at this point, you're either being insanely stubborn or just way way too dumb to be engaging in this conversation.

I'm not the poster you insulted in that thread. But do give more details on your method! It's entertaining.
Everything is explained well enough there that someone with basic stats knowledge and even average intelligence can understand that I'm 100% right. So if you don't get it, what does that say?
Can you give an example of a hypothesis test using a coefficient?
Yes. Already explained in detail in the other thread: For a regression, you compare the coefficient to a t distribution multiplied by the standard error. Second thing is still a distribution, of course. Voila, coefficient as a test statistic. Understand at long last?
The coefficient isn’t a test statistic there. The tstatistic is. The fact that you’re using algebra doesn’t change that...
Nope. Completely wrong. You're taking the coefficient and comparing it to a distribution. Hence, the coefficient is being used as a test statistic. Look at any definition of test statistic and it will qualify. I think you're coming around to the fact I'm right.Which distribution are you comparing the coefficient to?
Cool, this has only been mentioned a dozen times, so what's the harm in once more?
Normal test: b/se ~ t dist
Alternative: b ~ se*t distThe distribution in the alternative is the standard error times the t distribution. Look at what's going on there. We're taking a statistic (b), right? Now we're comparing it against a distribution, right? Then we're making conclusions and deducing p values to test a hypothesis, yeah? Well, that's what a test statistic is. QED.

No one is saying anything about her race. You keep bringing it up. The other posters are taking about her lack of knowledge of statistics, and the errors in her work.
That's hilarious. The posters are constantly mentioning her race and how she's the supposed beneficiary of AA. Some other trolls will come in and explain how Haitians have low IQs. So please don't play dumb there.
No one has identified any "error" whatsoever in a single publication of hers nor any lack of knowledge of statistics. This is something you've invented in your mind to satisfy your inferiority complex. The worst they've found is that CG used the completely standard interaction term techniques of the time and didn't use a method published two years after her 2004 paper. That's literally it.