## yashar806 2 years ago littile question on statistics, picture below

1. yashar806

2. yashar806

can you explain that?

3. yashar806

4. JuanitaM

yes let me see how I want to explain it

5. JamesJ

The second part is asking you for the value of R^2. Sounds familiar?

6. yashar806

ow ` I just have to square it ?

7. yashar806

what if it asks The Buick Regal has a horsepower of 127. What is the predicted mileage for the Buick Regal?

8. yashar806

i have to plug 127 into function?

9. JamesJ

yes

10. yashar806

what about this one Predicting the value of a response variable for an individual whose x-value is out of our range of data is called

11. JamesJ

Then you can still use the regression line to derive an predicted value for y, given the x. However, it is likely to be less reliable for an x outside the range of x in the original sample, particularly if it is a long way outside of that range.

12. yashar806

so how should I answer this question ?

13. JamesJ

I'm guessing just use your regression line if it asking you for an actual predicted y value

14. yashar806

Im sorry, I dont get it

15. yashar806

is it going to be extrapolation

16. JamesJ

Yes

17. yashar806

I think it asking for definition of extrapolation

18. JamesJ

Well, that's why you have a textbook. Look it up.

19. yashar806

so, my last question is What is the residual for the Chevy Malibu?

20. JamesJ

Residual is the difference between the predicted and actual value. So calculate the predicted value and compare it to the actual.

21. yashar806

So, I have to plug 251 and 28 into fuction?

22. JamesJ

No. You are predicting one variable in terms of the other, using the regression line. You are predicting the value of y, given x. You are predicting the value of mileage as a function of horsepower. That gives you the predicted value of mileage. The residual is then the difference between this predicted number and the actual mileage. These are fundamental concepts. I recommend you go back and re-read your lecture notes and/or the chapter in your text book to make sure you have this straight.

23. yashar806

ok , got it thanks

24. yashar806

I got the answer for that -2.1380

25. yashar806

could you expalin this one?

26. yashar806

Suppose we had instead measured mileage in kilometers per liter (1 m/g = 0.425 km/l). What would be the value of the correlation between X and Y?

27. phi

to change from mile per gallon to km per liter you would multiply each mileage value by a conversion factor (a fixed number). call that number a. If you plug in that factor everywhere into your formula how does it change the final correlation ?

28. yashar806

satys same?

29. yashar806

is it 0.8651 ?

30. phi

yes, it stays the same. But you can prove it to yourself.

31. yashar806

but when I enter the answer 0.8651 . it says , it's wrong

32. yashar806

now ,its right , thank you very much