Statistics II Spring 2023

Assignment (10)
Name: Student ID:
Name: Student ID: *
1. (25%) Suppose we want to know if the new pain killer is more effective, use the Wilcoxon rank sum test
to answer the following questions. The data for the two medicines are listed below:
The new pain killer 100 6 100 6 60 2 . 5 100 6
Aspirin 60 2 5
.
35 1 100 6 100 6
The responses were coded as follows: focaresoss ave the same
)Ho Thetiwpopulaion
100 = The drug was extremely effective.
e thenyht of the
60 = The drug was quite effective. H . iThe locaton ofthenew pansiller13
location of
dspmin
40 = The drug was somewhat effective. 返) T =L - . 5

5 TE 5 = D

* #
35 = The drug was slightly effective.
10 = The drug was not at all effective.
C) T == 18 σ÷= ☆= 。
·4
□≈ 34641

a. (5%) Write down H0 and H1 in this hypothesis testing. τ=T 2D 5 Z 州以 5

≈ n i
.
= -
= o .
2

464 )
b. (10%) Find the test statistics T1 and T2 . on 2| M new 20 . = 1645 ⽆照夠證登據説明 5
3 .

c. (10%) Determine whether we can infer with 5% significance level that the new painkiller is more pan lasle 8

effective than aspirin 較 aspmm 有效

⼭
、

2. (25%) A matched pairs experiment yielded the following results, and we want to know if the location of
population 1 differs from the location of population 2. Use the Wilcoxon signed rank sum test to answer
the following questions.
Pair 1 排有 2 5
.

2 6 3 7 4 2 -

5 5 25
.

6 2 5
-

7 5
Sample1 15 14 14 18 20 19 15
分盖 1 —
34 -
时 | L
Sample2 16 17 18 19 19 20 17
thesame locasns are ae Theto popuaitn
locasns atferent
Theto H .
popukaon 2

a. (5%) Write down H0 and H1 in this hypothesis testing. tb以

+
b. (10%) Find rank sums of positive difference T and negative difference T . − τ 25 T = 25 5
-

-
-

c. (10%) Can we infer at the 5% significance level that the location of population 1 differs from the
510
2
z 1
944 =
1 96
-

>
location of population 2?
-
-

85
-
σ "
= 14
不相等
=
@ EC T =
5 916 天路句證據明2 Sampe的 populaton
.

3. (25%) Use the Kuskal-Wallis test on the following data to conclude whether we can infer the population
locations differ
Sample1 25 8 20 4 5 -
15 时
Sample2 19 3 21 6 23 时

Sample3 27 18 映 20 火5

@2 of ifferx
Ho Trelocates all three ppuhktitmsarethesume Hi Aeleastem pupulaaonlocatr,
a. (5%) Write down H0 and H1 in this hypothesis testing.

v 3 + 5 + 5)
此 155
-516
b. (10%) Find the test statistic H for Kruskal-Wallis test. H =
(+ ) 11 -

3 <a
c. (10%) What is your conclusion at 5% significance level?

* (n )
5
Thores ot enough endene = 8
30 15
.

0
《此
- =

λ
-

0 05 , 2
.
= 5 qd o
reject 1 Ho
4. (25%) The following data were generated from a blocked experiment. Conduct a Friedman test to
determine whether at least two population locations differ.
Treatment
Block 1 2 3
1 7 久5 5 3 1 5

点
-

2 4 3 6 7 久5
3 8 q 6 5 -
5 3 1 5 .

[ ⑤ ⑤
a. (5%) Write down H0 and H1 in this hypothesis testing.
b. (10%) Find the test statistic F𝑟 for the Friedman test
c. (10%) What is your conclusion at 5% significance level?
@
Ho Trelocaons of all threo
popukeions thesarme
ane
s
zy
—
- 152 (
3 u 5 +
+10 - 5) -3
HAeleastthropopnlationlo catroms 我fferx
⼆
36238
5 .
-
=
232 5 -

c)
≥ 0 .
05 ,
2
= q9
5
< 232 5 -

We an reject Ho
At leastthepopbutitnlocaxeitns ditfer
#