00:01
So here we have a simple linear regression model.
00:03
Our predicted value for y, which is the price of the house, is 33 ,478, plus 62 .4 times x, right? we are also told the coefficient of correlation r is equal to 0 .63.
00:19
So a, we want to predict y hat if x is equal to 1860.
00:26
Well, all we need to do is plug in for y hat.
00:31
3478 plus 62 .4 times 1860.
00:37
Of course this is something that you are forced to put in, i mean, maybe you can do this in your head, but for me that would be basically impossible.
00:48
And we get a predicted selling price of $149 ,000, $542.
00:56
So that would be the predicted selling price.
00:58
We are now but we observe y is equal to instead $165 ,000.
01:11
So these are not equal.
01:14
And b, they could be not equal for a variety of reasons, right? the first one is simply randomness.
01:22
Remember, this model is not attempting to predict every single observation perfectly.
01:27
It's only predicting the average, right? it's a model for the expected value.
01:33
It's not a model...