This exercise is once again looking at S 80 schools. But this time looking at how we can use an S A T to predict a GPS and a late gentile people could perform a college level. So part a asked us to write the equation of the regression light. Now, everything we need for this is given in the question. So we are predicting our g p a. They want g p A. Remember the hot to say it's a prediction predicted g p A is now are a value is given by the intercept Kerr fish and the question so minus 1.26 to remember the minus side and then our s a T coefficient is again given to us for the positive. No point no to no, not 21 for that city. Uh, that is old neat moving on pop be asked us to explain what the Y intercept of this Russian line indicates. Now Warren stepped is given to us hit now Usually in this imminent in any context, the won't set would represent predicted G p a. When s a T equals zero. Um, but it's hard for that toe kind of make any sense in this context because we've got a negative number as the intercept. But it does indicate a predicted debate when a city is zero and then looking at C l sister interpret this slope, which is this done? So what this is saying is for inch one point set in groups, we would predict E g p a. Increase off that value of the slopes off. No point, no, no to want full quints. That's what that means. So we will probably miss asking us to predict the GPS. A freshman who scored a combined 21 hundreds on This is just substituting into the equation we worked out in part A. So I predicted GDP is one point minus 1.262 plus no point, no North 214 as a very substrate in that 2100 these two moles plan together, come to four point for no for So if we have negative one point 262 that's 4.494 that is gonna come to three point to three to So the answer point for Part D and finally looking at Part E on it not finally looking at high, it says, Based on the statistics, How effective do we think SAT schools would be in predicting academic except success during the first most of the freshman year up to college? So the best indicator of this a one of the best and kids to this is our r squared value, which we given as 22.1%. Now it's showing that 22.1% of the variability in GP A schools in the first Master can be explained with this model so it could be explained with the S A T schools. Why was no high? It says it can give some indication. So I'd say Molterer indication. So while it shouldn't be used to fully gauge what people are gonna get, it can give you some kind of idea. But yeah, 80 well, 78.9% of very billeting do free schools could be explained with other things, so it definitely shouldn't be taken as everything that goes towards GP a school now in the last five, it says. As a student, if one it says as a student, would you rather have a positive or negative residual in this context. So? So that s manager about the or residual is Ah, actual or observe school minus are predicted school. Now, I did a We'd want the actual school three o'clock as possible. So for if our actual scores higher than I predicted, in other words, we did better than they thought. That would be groups and indigent that Gregson zero. So we'll be looking a positive residual. And I was a student. You definitely will, Max, if you get a negative residual.