LOL
BREAKING: CG@y to step down from administrative role to focus on her "research"

The worst affirmative action leech in the history of US academia.
Wait, is this the same CG with 5 *solo* top3s? Because that's an awfully uncomfortable fact!Those articles are as intellectually honest as Lacour's Science article.
And yet when challenged to find anything wrong with them, you can't! Just distract and blather. She's a winner. You're a loser. All that matters.

It's been common knowledge for years and pointed out multiple times on this forum. I will enumerate just a few off the top of my head. Logically inconsistent assumptions in ecological inference models, incorrect specification and interpretation of interaction terms in regression analyses, unreliable estimates due to issues such as models with high multicollinearity, incorrect derivations of predicted probabilities, no computation of confidence intervals around estimates to determine whether various subgroups are statistically distinguishable from each other, lack of replicability etc. Go ahead and google polmeth papers by Shotts & co. for more detailed analyses.
The worst affirmative action leech in the history of US academia.
Wait, is this the same CG with 5 *solo* top3s? Because that's an awfully uncomfortable fact!
Those articles are as intellectually honest as Lacour's Science article.And yet when challenged to find anything wrong with them, you can't! Just distract and blather. She's a winner. You're a loser. All that matters.

It's been common knowledge for years and pointed out multiple times on this forum. I will enumerate just a few off the top of my head. Logically inconsistent assumptions in ecological inference models, incorrect specification and interpretation of interaction terms in regression analyses, unreliable estimates due to issues such as models with high multicollinearity, incorrect derivations of predicted probabilities, no computation of confidence intervals around estimates to determine whether various subgroups are statistically distinguishable from each other, lack of replicability etc. Go ahead and google polmeth papers by Shotts & co. for more detailed analyses.
The worst affirmative action leech in the history of US academia.
Wait, is this the same CG with 5 *solo* top3s? Because that's an awfully uncomfortable fact!
Those articles are as intellectually honest as Lacour's Science article.
And yet when challenged to find anything wrong with them, you can't! Just distract and blather. She's a winner. You're a loser. All that matters.Hahaha, this doofus is so clueless he thinks multicollinearity causes bias. Note zero support for a single claim, other than vague reference to an unpublished conference paper. False claim about lack of replicability (all data is from widely available datasets). False claim about interaction terms. Etc. Just nothing.
Again, this is a pure loser, aggrieved about a black woman who has published 5 solo top3s and is a dean at Harvard. She's a winner. He has produced nothing, so feels the need to lie and snarl about other's work to hide this. Sad!

^ Go the the library, grab a copy of W. Greene's Econometric Analysis, 7th edition, and look at p. 130. It says that a consequence of multicollinearity is that coefficients may have the incorrect sign or implausible magnitudes. You also seem confused about the definition of unbiasedness.

Don't be disingenuous. Where is the replication code?
False claim about lack of replicability (all data is from widely available datasets).People who are smart enough to read these papers and understand what multicollinearity is, can easily replicate the papers. Sorry that you need a walkthrough explaining how to run a regression.

Want to know the clearest sign that you're an incredibly strong empirical researcher? If a jealous critic is asked to explain flaws in your work and the first thing they say is "uhhh, I think there's some, uh, multicollinearity in, um, some models, maybe." That means you're a superstar. Maybe Harvard Dean material!!!

Don't be disingenuous. Where is the replication code?
False claim about lack of replicability (all data is from widely available datasets).People who are smart enough to read these papers and understand what multicollinearity is, can easily replicate the papers. Sorry that you need a walkthrough explaining how to run a regression.
There are things beyond the simple regression that she is not transparent about, or worse, she contradicts herself. Let's take the 2004 APSR an an example. I'll give you two quotes from that:
“The simulations in Figure 2 also show that African Americans are more likely to express concern about the persistence of antiblack discrimination” (p. 555)
“The probabilities are derived from the ordered probit estimates reported in Table 1, Model 1. “ (p. 556, referring to Fig. 2)
So which one is it? Is it a simulation? And if so, which software was used, which variables were held at which values? Or is it a derivation, and if so, which formula is applied, and how was it derived? Where in the paper are the relevant equations?
If you think it's so clear and easy to replicate without code, go ahead and answer the questions above.

Don't be disingenuous. Where is the replication code?
False claim about lack of replicability (all data is from widely available datasets).
People who are smart enough to read these papers and understand what multicollinearity is, can easily replicate the papers. Sorry that you need a walkthrough explaining how to run a regression.There are things beyond the simple regression that she is not transparent about, or worse, she contradicts herself. Let's take the 2004 APSR an an example. I'll give you two quotes from that:
“The simulations in Figure 2 also show that African Americans are more likely to express concern about the persistence of antiblack discrimination” (p. 555)
“The probabilities are derived from the ordered probit estimates reported in Table 1, Model 1. “ (p. 556, referring to Fig. 2)
So which one is it? Is it a simulation? And if so, which software was used, which variables were held at which values? Or is it a derivation, and if so, which formula is applied, and how was it derived? Where in the paper are the relevant equations?
If you think it's so clear and easy to replicate without code, go ahead and answer the questions above.Hahahahahahahahahaha! Hahahahahahahahahaha! Hahahahahahahahahaha!
That's the best you got!!! You're not familiar with "simulations" being used to refer to results of estimates. I guess you haven't read many papers. This is hilarious, cringey stuff! I love it.

Or is it a derivation, and if so, which formula is applied, and how was it derived? Where in the paper are the relevant equations?
Donat: Please, I don't understand what ordered orbit is! You need to give me the equations or I'll never figure it out! What's a stata! What do you mean! How do numbers work!
Hahaha, this goof reminds me of Zoolander. The files are *in* the computer?!