Scribd
Upload
40 views

Homework2 Chapter4 Solution

This document provides solutions to 4 homework problems. Problem 1 analyzes diameter data and rejects the null hypothesis. Problem 2 finds no evidence of difference between technicians' measurements. Problem 3 compares means of 2 groups and finds no significant difference. Problem 4 uses a boxplot and t-test to determine smoking mothers have babies with significantly lower weight than non-smoking mothers.

Uploaded by

nc
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
40 views

Homework2 Chapter4 Solution

This document provides solutions to 4 homework problems. Problem 1 analyzes diameter data and rejects the null hypothesis. Problem 2 finds no evidence of difference between technicians' measurements. Problem 3 compares means of 2 groups and finds no significant difference. Problem 4 uses a boxplot and t-test to determine smoking mothers have babies with significantly lower weight than non-smoking mothers.

Uploaded by

nc
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

Homework 2 Solution

Problem 1 (24 points)

Solution:
(a) (10 Points)

𝐻! : 𝜇 = 8.25 𝑣𝑠. 𝐻" : 𝜇 ≠ 8.25. Reject 𝐻! if |𝑍! | > 𝑍#/%

Conclusion: It is concluded that we should reject the null hypothesis that the mean diameter
is 8.25cm.

(b) (4 points)
Z value 6.78 is out of the range in the table, therefore, P-value is close to 0, or smaller
than 0.00006.

(c) (10 points)

Problem 2 (24 points)

Solution:
(a) 8 points
NOTE: Correction à |Z0| = |-2.2667| = 2.2667 < Zα/2
(b) 8 points

(c) 8 points

&√(
The reason that we can approximate the first term Φ 12.58 − ".%
3 to 1 is that:

−√𝑛 1
Φ 52.58 − 8 ≥ Φ :2.58 + < = Φ(3.41) = 0.9997 ≈ 1
1.2 1.2

Problem 3 Solution: (32 points)

(a) (10 pts)


Test H! : µ" − 𝜇# = 0 vs H" : µ" − 𝜇# ≠ 0. Reject H! if |t !| > 𝑡!,%# &%" '# .
"
(𝑛" − 1)𝑠"# + (𝑛# − 1)𝑠## (7 − 1)0.115# + (8 − 1)0.11#
s( = . =. = 0.112
𝑛" + 𝑛# − 2 7+8−2
)̅# ')̅ " "../.'"..//
Therefore t ! = # #
= # #
= −0.086 , 𝑡!,%# &%" '# = 𝑡!.!#0,". = 2.160
+$ ,% &% !.""#,&&' "
# "

Do not reject H! , and conclude that there is no sufficient evidence of a difference


between measurements obtained by two technicians.
(b) (4 pts)

The practical implication of this test is that we don’t have enough evidence/data to
tell the difference between the two technicians. If we collect more data, it is
possible that we will end up rejecting the Null Hypothesis, which implies the
difference between the two technicians is significant by the data we collected.
(c) (10 pts)
n1 = 7, x1 = 1.383, S1 = 0.115; n2 = 8, x2 = 1.388, S2 = 0.11

a = 0.05, ta/2, n1+n2-2 = t0.025, 13 = 2.1604


(𝑛" − 1)𝑠"# + (𝑛# − 1)𝑠## (7 − 1)0.115# + (8 − 1)0.11#
s( = . = . = 0.112
𝑛" + 𝑛# − 2 7+8−2

1 1
(1.383 − 1.388) − 2.1604(0.112), + ≤ (𝜇! − 𝜇" )
7 8

1 1
≤ (1.383 − 1.388) + 2.1604(0.112), +
7 8

−0.13 ≤ (𝜇! − 𝜇" ) ≤ 0.12

(d) (8 pts)
The JMP procedure is given:

1) First, put the data into a table.


2) Analyze -> Fit y by x
3) Red arrow -> Means/ANOVA/Pooled t
The result is given as follows:
Oneway Analysis of Data By Label

Oneway Anova
Summary of Fit

Rsquare 0.000491
Adj Rsquare -0.07639
Root Mean Square Error 0.112213
Mean of Response 1.385333
Observations (or Sum Wgts) 15

t Test
y-x
Assuming equal variances

Difference 0.00464 t Ratio 0.079945


Std Err Dif 0.05808 DF 13
Upper CL Dif 0.13011 Prob > |t| 0.9375
Lower CL Dif -0.12082 Prob > t 0.4687
Confidence 0.95 Prob < t 0.5313

Analysis of Variance

Source DF Sum of Mean Square F Ratio Prob > F


Squares
Label 1 0.00008048 0.000080 0.0064 0.9375
Error 13 0.16369286 0.012592
C. Total 14 0.16377333

Means for Oneway Anova

Level Number Mean Std Error Lower 95% Upper 95%


x 7 1.38286 0.04241 1.2912 1.4745
Level Number Mean Std Error Lower 95% Upper 95%
y 8 1.38750 0.03967 1.3018 1.4732

Std Error uses a pooled estimate of error variance


Missing Rows 1

Problem 4 Solution: (20 points)


a) (5 points) For the plot, you should use the boxplot to compare two groups.
The procedure is given as follows:
• Graph -> Graph Builder

• Drag bwt to y-axis

• Drag smoke to x-axis

• Select Boxplot

b) (10 points) Hypothesis Test:


The Null hypothesis is that the two groups (whether smoke or not) will have
the same weight.
The alternative hypothesis is that the two groups (whether smoke or not) will
have different weight.
Here is the procedure to conduct the hypothesis test.
• Make sure that smoke is an ordinal variable. If not, please change as the
following

• Analyze -> Fit y by x

• Drag bwt to y

• Drag smoke to x and OK, you should have this

• Click the red arrow and select t Test and you will have the following results
Therefore, we can conclude that the test is significant by the following two information:
1) 0 is outside the Upper CL Dif and Lower CL Dif. Or 0 ∉ [−488.98, −78,57]
2) The P-value for two-sided test 0.007 < 0.05.

Overall, our data does support that the weight between the smoking and non-smoking mother
would be different.
c) (5 points) The p-value is 0.007 for a two-sided test.

You might also like