Bill Cassidy actually said...

"Is there a second variable you would throw in there? Because that was not correlated with the teacher."

"Now, with teachers, what are the top two variables?"

"Again, you find some improvement in mathematics, but otherwise, these highly motivated kids with great verbal skills are still little able to impact kids who are otherwise anchored down by a terrible ..."

"So you must have regression analysis within that. Can you give me your top four, actually give me the top two."