Academic Integrity: tutoring, explanations, and feedback — we don’t complete graded work or submit on a student’s behalf.

EXERCISE #1 Small Is Beautiful: The relationship between Height and Longevity. T

ID: 384548 • Letter: E

Question

EXERCISE #1 Small Is Beautiful: The relationship between Height and Longevity. The data below shows the President of USA, his age at time of death, and his height. Age at Time Height President of Death (inches) Madison 85 64 Van Buren 79 66 B. Harrison 67 66 J, Adams 90 67 J.Q. Adams 80 67 Jackson 78 73 Washington 67 74 Arthur 56 74 F. Roosevelt 63 74 L. Johnson 64 74 Jefferson 83 74.5. A simple linear regression of the model: LONGEVITY = b + b HEIGHT was run and the computer output results are shown below. Use the output to respond to the following questions:

(a) What sort of relationship exists between the height of the president and the longevity of the president?

(b) Interpret the meaning behind the values of the estimated b and b .

(c) Predict the expected longevity of a president who is: (i) 48 inches, (ii) 60 inches, (iii) 70 inches, (iv) 80 inches tall.

(d) At a level of significance, = 0.05, test for the significance of the relationship between height and longevity. What is your conclusion? What comments do you have on the predictions you made in part (c)?

(e) Give an interpretation to the R-Square value in the computer output.

Explanation / Answer

LONGEVITY = 174.424 - 1.430722 HEIGHT

  R-Squared                                = 0.309636

    Adjusted R-Squared = 0.232929

    Standard error of estimate = 9.504369

    Number of cases used = 11

Analysis of Variance

                                                                                           p-value

    Source             SS df MS F Value Sig Prob

    Regression    364.63900 1 364.63900 4.03661              0.075434

    Residual        812.99730 9     90.33304

    Total            1177.63600 10

LONGEVITY AND ONE'S HEIGHT

REGRESSION COEFFICIENTS FOR LONGEVITY

                                                                                  Two-Sided p-value

    Variable       Coefficient            Std Error t Value Sig Prob

    Constant      174.42400            50.15620      3.47761 0.006965

    HEIGHT         -1.43072 0.71211       -2.00913              0.075434 *

    Standard error of estimate          = 9.504369

    Durbin-Watson statistic   = 1.900782

ANSWER

Model is                         LONGEVITY = b + bHEIGHT

Estimated regression is: LONGEVITY = 174.424 - 1.430722 HEIGHT

         

b = - 1.430722 is negative, therefore relationship is INVERSE

ANSWER   b = 174.424 is the intercept = expected longevity of a president of zero height

b = - 1.430722 the slope, for every one inch increment in height, longevity drops by 1.430722 years

(c)

ANSWERS

Estimated regression is: LONGEVITY = 174.424 - 1.430722 HEIGHT

(d)  ANSWERS Model is LONGEVITY = b + bHEIGHT

H: b= 0 is the null hypothesis, and

H: b 0 is the alternative hypothesis

Using the F-test

Analysis of Variance

                                                                                           p-value

    Source             SS           df         MS           F Value     Sig Prob

    Regression    364.63900   1    364.63900      4.03661              0.075434

    Residual        812.99730             9     90.33304

    Total            1177.63600         10

The p-value = 0.075434 > = 0.05

DO NOT REJECT NULL:

H: b= 0 is the null hypothesis

Therefore Null is correct; the two variables are not related

USING THE t-TEST

LONGEVITY AND ONE'S HEIGHT

REGRESSION COEFFICIENTS FOR LONGEVITY

                                                                                  Two-Sided p-value

    Variable       Coefficient            Std Error      t Value     Sig Prob

    Constant      174.42400            50.15620      3.47761    0.006965

    HEIGHT         -1.43072           0.71211       -2.00913              0.075434 *

p-value = 0.075434 * > = 0.05

The p-value = 0.075434 > = 0.05

DO NOT REJECT NULL:

H: b= 0 is the null hypothesis

Therefore Null is correct; the two variables are not related

What comments do you have on the predictions you made in part (c)?

Since the the relationship between the two variables is not statistically significant, our predictions in part (c) are not reliable.

(e) ANSWER

              R-Squared        = 0.309636 = 30.9636%

Height accounts for 30.9636% of variation in longevity