Scribd
Upload
134 views

Assignment No.6

This document contains a multiple choice assignment with questions related to statistics and data analysis. Question 1 provides a regression equation relating advertising spending (x) to gross sales (y) for electronics companies. Question 2 displays height and IQ score data for 8 high school girls and asks about the correlation. Question 3 asks about calculating and interpreting the coefficient of determination from a given correlation coefficient. Question 4 asks to estimate a regression line from the height-IQ data and verify an identity related to total, regression, and residual sum of squares.

Uploaded by

Minza
Copyright
© © All Rights Reserved
Available Formats
Download as XLSX, PDF, TXT or read online on Scribd
134 views

Assignment No.6

This document contains a multiple choice assignment with questions related to statistics and data analysis. Question 1 provides a regression equation relating advertising spending (x) to gross sales (y) for electronics companies. Question 2 displays height and IQ score data for 8 high school girls and asks about the correlation. Question 3 asks about calculating and interpreting the coefficient of determination from a given correlation coefficient. Question 4 asks to estimate a regression line from the height-IQ data and verify an identity related to total, regression, and residual sum of squares.

Uploaded by

Minza
Copyright
© © All Rights Reserved
Available Formats
Download as XLSX, PDF, TXT or read online on Scribd
You are on page 1/ 8

ASSIGNMENT NO.

6
Q1: A researcher took a sample of 25 electronics companies and found the following relationship between x and y where x i
of dollars) spent on advertising by a company in 1996 and y represents the total gross sales (in millions of dollars) of that co
ŷ = 3.4 + 11.55x is the least-squares regression line.

a) An electronics company spent $2 million on advertising in 1996. What are the expected gross sales for 1996?

ŷ = 3.4 + 11.55x q
ŷ 26.5

The gross sale of company is $ 26.5 millions in 1996.

b) Suppose four electronics companies spent $2 million each on advertising in 1996. Do you expect these four companies to
gross sales for 1996? Explain.

Answer: No, gross sales of companies may be different due to randomness


ip between x and y where x is the amount of money (in millions
millions of dollars) of that company in 1996.

s sales for 1996?

pect these four companies to have the same actual


Q2: The height (in inches) of 8 high school girls and their scores on an IQ test is given below

a. display the data in scatter plot


Answer: Scatter plot which is plotted below shows positive relationship, as height will increase iq score will also increase.

b. Describe the type of correlation, and interpret the correlation in context of the data.
Solution:

Height (x) IQ score (y) (x-x_bar) (y-y_bar) (x-x_bar)(y-y_bar) (x-x_bar)^2 (y-y_bar)^2


62 109 -0.125 -1.25 0.15625 0.015625 1.5625
58 102 -4.125 -8.25 34.03125 17.015625 68.0625
65 107 2.875 -3.25 -9.34375 8.265625 10.5625
67 114 4.875 3.75 18.28125 23.765625 14.0625
59 96 -3.125 -14.25 44.53125 9.765625 203.0625
64 110 1.875 -0.25 -0.46875 3.515625 0.0625
65 116 2.875 5.75 16.53125 8.265625 33.0625
57 128 -5.125 17.75 -90.96875 26.265625 315.0625
Mean 62.125 110.25 12.75 96.875 645.5

r 0.050986618

Interpretation: As correlation coefficient is close to zero it implies that the two variables height and iq scores
have little to no linear relationship or dependence.

IQ score (y)
140
120
100
80
60
40
20
0
56 58 60 62 64 66 68
ore will also increase.
Q3: Use the value of the correlation coefficient r to calculate the coefficient of determination r2. What does this tell you a
explained variation of the data about the regression line?
Solution:

a.
r 0.465
r^2 22%

That is 21% of the variability of y is explained frm the variation of x in regression line

b.
r -0.957
r^2 92%

That is 92% of the variability of y is explained frm the variation of x in regression line
tion r2. What does this tell you about the
ine?
Q4: Use the data of Q2:
a) Estimate a regression line for IQ Score, y on height of high school girls, and interpret the meaning of slope and intercept.
b) Verify the identity SST = SSR + SSE

Solution:
a.
Height (x) IQ score (y) (x-x_bar) (y-y)bar) (x-x)bar)^2 (x-x_bar)*(y-y_bar)
62 109 -0.125 -1.25 0.015625 0.15625
58 102 -4.125 -8.25 17.015625 34.03125
65 107 2.875 -3.25 8.265625 -9.34375
67 114 4.875 3.75 23.765625 18.28125
59 96 -3.125 -14.25 9.765625 44.53125
64 110 1.875 -0.25 3.515625 -0.46875
65 116 2.875 5.75 8.265625 16.53125
57 128 -5.125 17.75 26.265625 -90.96875
Mean 62.125 110.25 0 0 96.875 12.75

Bo 102.07 intercept
B1 0.131612903 slope
y= 102.07+0.1316129x
Interpretation: If height of school girls increases by 1% then iq score will increase by 0.1316129

SST = SSR + SSE


97886.0000 97242.1781 643.8219

140

120

f(x) = 0.131612903225
100 R² = 0.0025996351915

80

60

40

20

0
56 58
the meaning of slope and intercept.

y_predicted Error SSE SSR SST


110.23 -1.23 1.5216 12151.4352 11881
109.71 -7.71 59.3993 12035.6471 10404
110.63 -3.63 13.1652 12238.6400 11449
110.89 3.11 9.6621 12296.9498 12996
109.84 -13.84 191.5099 12064.5421 9216
110.50 -0.50 0.2468 12209.5371 12100
110.63 5.37 28.8542 12238.6400 13456
109.58 18.42 339.4628 12006.7867 16384
643.8219 97242.1781 97886.0000

SST 97886.0000

IQ score (y)
140

120

f(x) = 0.131612903225807 x + 102.073548387097


100 R² = 0.002599635191525

80

60

40

20

0
56 58 60 62 64 66 68

You might also like