Sei sulla pagina 1di 4

PADP 7120: PS 6

Solutions

1. This table presents sample data for the number of hours spent by individual students
outside of class on a course on statistics during a period of 3 weeks and their scores on
an examination given at the end of the period.
(a) Plot the data.
(b) Determine the regression equation for predicting the examination grade given the
number of hours spent on the course and enter the regression line on the scatter
diagram.
(c) Construct the 95 percent confidence interval for the examination score given that
the student devoted 30 hours to the course preparation, using the standard error
of the estimate as the measure of uncertainty.
(d) Construct it for the mean grade.
(e) Test the null hypothesis that the slope of the regression line is zero. Use the 1
percent level of significance. Interpret.
(f) Write 50 ± 10 words interpreting what you found in answering this problem.
Sampled Student 1 2 3 4 5 6 7 8
Hours of Study (X) 20 16 34 23 27 32 18 22
Exam Grade (Y) 64 61 84 70 88 92 72 77
Solution:
We want the solutions a and b for the line Y = a + bX. Start with b, the slope.
The formula for finding the slope of the line is:

SXY
b =
SXX
P
XY − nX̄ Ȳ
b = P 2
X − nX̄ 2

1
Build a table:
Student X Y XY X2 Y2
1 20 64 1280 400 4096
2 16 61 976 256 3721
3 34 84 2856 1156 7056
4 23 70 1610 529 4900
5 27 88 2376 729 7744
6 32 92 2944 1024 8464
7 18 72 1296 324 5184
8 22 77 1694 484 5929
Total 192 608 15032 4902 47094
Mean 24.00 76.00
P
XY − nX̄ Ȳ
b = P 2
X − nX̄ 2
15032 − 8(24)(76)
=
4902 − 8(24)2
440
=
294
= 1.497

Recall that a = Ȳ − bX̄

a = Ȳ − bX̄
= 76 − 1.497(24)
= 40.082

We estimate the regression equation: Y = 40.08 + 1.50X


Suppose that a student studied 30 hours. Here’s how we calculate the 95 percent
confidence interval for our estimate of that student’s grade.
Our best guess about that student’s grade is:

Ȳx = a + bX
= 40.082 + 1.497(30)
= 84.980

2
We need the standard error of the estimate if we want to calculate the 95 percent CI:

rP P P
Y2−a Y − b XY
sY.X =
n−2
r
47094 − 40.082(608) − 1.497(15032)
=
8−2
r
227.497
=
6

= 37.916
= 6.158

We also need a tcrit value:


df = n − 2 = 6, α = 0.95
The 95 percent CI is:

Ȳx ± tsY.X
84.980 ± 2.447(6.158)

The upper CI is 100.05. The lower CI is 69.91.


Suppose that a student studied the mean number of hours. Here’s how we calculate
the 95 percent confidence interval for our estimate of that student’s grade.
Our best guess about that student’s grade is:

ȲX̄ = a + bX
= 40.082 + 1.497(24)
= 76.00

The 95 percent CI is:

ȲX̄ ± tsY.X
76.00 ± 2.447(6.158)

3
The upper CI is 91.07. The lower CI is 60.93.
Now test the null hypothesis that the slope is zero. We’ll use the 1 percent significance
level.

sY.X
sb = pP
X 2 − nX̄ 2
6.158
=
17.146
= 0.359

b − β0
t =
sb
1.497 − 0
=
0.359
= 4.17

tcrit = 3.707
Is t > tcrit ?
Yes, therefore reject the null.

Potrebbero piacerti anche