3) Now let's assume you wanted to predict weights from heights. In other words, now let's use the same data from problem (1) and do a regression instead. (a) Calculate bo and bị. (b) Carefully draw...

question 4

Extracted text: 3) Now let's assume you wanted to predict weights from heights. In other words, now let's use the same data from problem (1) and do a regression instead. (a) Calculate bo and bị. (b) Carefully draw your least square regression line on the plot you made in 1(a). (Don't just “sketch", be a little careful). 4) (a) Now do a significance test of Ho: B1 = 0. Show all your calculations (including your residual calculations). Again, note that this should be a one sided test (why?). (b) Compare your t' from 4(a) with your i' from problem 2. Are they the same? This is not a coincidence, although once you do more complicated analyses, you can't rely on this "equivalence". (Problems 5 - 7 on next page)


Extracted text: 1) You compare the height (cm) and weight (kg) of 5 adult women. You get the following results: height 152.4 170,2 157.5 177.8 167.6 weight 54.0 64.9 59.4 70.3 61.7 (a) Construct a scatter plot of height (x) vs. weight (y) (b) Calculate the following (show your work for SSep): S, SS, y (c) Calculate the correlation coefficient (r) 2) Use the information from problem 1. Perform a complete test of the hypothesis that the population correlation coefficient (p) is 0. Show all steps (note - obviously this should be a one sided test! Make sure you know why!). 3) Now let's assume you wanted to predict weights from heights. In other words, now let's use the same data from problem (1) and do a regression instead. (a) Calculate bo and b. (b) Carefully draw your least square regression line on the plot you made in 1(a). (Don't just “sketch", be a little careful).