The problems with the linear probability model - part 2

[@LordOfNoobstown]

I am wondering, when ben factors at 7:00, does he see the factorization immediately with no algebraic manipulation or was their some auxiliary calculation that he left out for this video? I must say, I had to do at least three lines of algebra before i came up with this factorization. I hope ben replies :)

[@nathanbarnard7896]

at 7:49 how does the factorisation work? I get 1-2Bx

[@lingkiwong2386]

(-BXi)^2 is equivalent to (BXi)^2 as the outcome is positive anyways. When you factorize the (-BXi)^2 and take out one (BXi), the remaining part is (BXi) instead of (-BXi). I hope this helps.

[@jana19993]

yeah i'm also not sure if you factorised right, surely you get negative beta xi?

[@davinpaharia6265]

+Jan Ahmet (-beta xi)^2 = (betaxi)^2 this is what happens according to me. or else there is no valid explanation.

[@lastua8562]

@@davinpaharia6265 Right, thought the same. Any ideas on whether this is - purely mathematically - acceptable to do? (Though for our purposes it will be the same)

[@FaizShakirNure]

Thanks so much

[@brianclark4796]

This is counterintuitive. Why would error not be 0 or 1? If you have a false positive or false negative, error is 1. Otherwise, the difference between both model and estimation is 0 (b/c 0-0=0 and 1-1=0)

[@ronakpol1580]

Ben Lambert U did a mistake while factorising at the 7.12 min.. please check if its your mistake i am doing something wrong

[@lastua8562]

When you say there is a mistake you should clarify. I thought the same, but the squares of pos. and neg. values are the same, that is why it is getting the same result (purely mathematically, I am not sure whether it would be acceptable to factorise like this though, happy if someone can give further insight below)