00:03
They give us, is this a bunch of true and false problems? so they're pretty short.
00:08
So let's say here.
00:11
Actually, maybe i'll just do them all together and look through them all.
00:17
So they give us this y equals 3 .29x minus 4 .217.
00:25
And they ask us, is that have a negative correlation? and that's false because we can see here the slope here, and that's basically what tells us what the correlation is.
00:37
The slope is positive, which means that y is increasing as x increases.
00:44
And so when you have y increasing as x increasing, that's when you have a positive correlation.
00:50
So it's not a negative correlation.
00:52
We have positive correlation, and so the statement there is false.
00:58
Then the next one, they basically give us another linear model, and it is minus 0 .0.
01:04
0 .238x plus 25.
01:07
And they ask, does that have a negative correlation? and indeed, it does.
01:11
So that is true.
01:13
And that's because the slope here is negative.
01:16
So we can see that y decreases as x increases.
01:21
You know, it obviously, this offset is here, but that's, we don't really care about this term for the correlation.
01:27
We know that the correlation, the line goes something like this.
01:31
It's shifted up, but it still has a negative.
01:34
Slope.
01:35
So the correlation is negative and so that statement is true.
01:42
Now they ask us, they tell us that the correlation coefficient of some, you know, data and a model is minus 0 .9871.
01:57
And does that imply a good fit? well, it does imply a good fit because whenever r, the absolute value of r is close to one, that by definition, how r is defined, it means that the model passes through the data quite well.
02:18
So this is, the aptitude value of one of this is close to one.
02:23
So the model is a good fit.
02:25
Now, you know, this isn't a great fit, even though it seems like this is very close to one, because really the fit, you really kind of want, you know, at least something like this for a good fit.
02:38
So again, just because it looks really close to one, doesn't mean...