Loading documents preview...
CHI-SQUARE Prepared by: ANTONIO, RUSSEL N. BARADI, ARNEL F. LAZO, IVY U.
1
CHI-SQUARE
A chi-square (Οπ ) statistic is a test that measures how expectations compare to actual observed data.
2
CHI-SQUARE TEST
It is used to determine whether there is a significant difference between the expected frequencies and the observed frequencies in one or more categories. 3
How to Calculate the Chi-Square Statistic (Οπ )
1.
2. 3.
For each observed number in the table subtract the corresponding expected number π β πΈ. Square the difference π β πΈ 2 Divide the squares obtained for each cell in the table by the expected number for that cell
πβπΈ 2 πΈ
4
CHI-SQUARE TESTS
GOODNESS OF FIT TEST
TEST FOR INDEPENDENCE
TEST FOR HOMOGENEITY
- used to test whether a frequency distribution obtained experimentally fits an βexpectedβ frequency distribution that is based on the theoretical or previously known probability of each outcome.
- used to test whether or not two variables are independent.
- used to draw conclusion whether two populations have the same distribution
5
GOODNESS OF FIT TEST 6
GOODNESS OF FIT
β’ π»π : The population frequencies are equal to the expected
β’ β’ β’
frequencies. π»π : The null hypothesis is false πΌ: level of significance πππππππ ππ πππππππ: π β 1 πΆβπ¬ π π¬
β’ =Ο β’ From πΌ and π β 1, a critical value is determined from the chi-square Οπ
table.
β’ Reject π»π if Οπ is larger than the critical value (right-tailed test)
7
GOODNESS-OF-FIT TEST
1. Researchers have conducted a survey of 1600 coffee drinkers asking how much coffee they drink in order to confirm previous studies. Previous studies have indicated that 72% of Americans drink coffee. At πΌ = 0.05 , is there enough evidence to conclude that the distributions are the same?
Paytonbinnings.com 8
Researchers have conducted a survey of 1600 coffee drinkers asking how much coffee they drink in order to confirm previous studies. Previous studies have indicated that 72% of Americans drink coffee. At πΌ = 0.05, is there enough evidence to conclude that the distributions are the same?
GOODNESS-OF-FIT TEST
Response
Frequency
Response
Frequency
2 cups/week
15%
2 cups/week
206
1 cup/week
13%
1 cup/week
193
1 cup/day
27%
1 cup/day
462
2+cups/day
45%
2+cups/day
739
9
GOODNESS-OF-FIT TEST
Researchers have conducted a survey of 1600 coffee drinkers asking how much coffee they drink in order to confirm previous studies. Previous studies have indicated that 72% of Americans drink coffee. At πΌ = 0.05, is there enough evidence to conclude that the distributions are the same?
π»π :The population frequencies are equal to the expected frequencies. π»π : The null is false. πΌ = 0.05 πππππππ ππ πππππππ: π β 1 = 4 β 1 = 3 10
Researchers have conducted a survey of 1600 coffee drinkers asking how much coffee they drink in order to confirm previous studies. Previous studies have indicated that 72% of Americans drink coffee. At πΌ = 0.05, is there enough evidence to conclude that the distributions are the same?
GOODNESS-OF-FIT TEST
Response 2 cups/week 1 cup/week 1 cup/day 2+ cups/day
%of Coffee Drinkers 15% 13% 27% 45% Οπ ΰ·
E
O
O-E
πΆβπ¬
240 208 432 720
206 193 462 739
-34 -15 30 19
1156 225 900 361
πΆβπ¬ π¬
π
πΆβπ¬ π¬
π
4.817 1.082 2.083 0.5014
π
= π. πππ 11
Οπ = π. πππ
Percentage Points of the Chi-Square Distribution
0.50
0.25
0.10
0.05
0.01
1
0.455
1.32
2.71
3.84
6.63
2
1.386
2.77
4.61
5.99
9.21
3
2.366
4.11
6.25
7.81
11.34
4
3.357
5.39
7.78
9.49
13.28
5
4.351
6.63
9.24
11.07
15.09
6
5.348
7.84
10.64
12.59
16.81
7
6.346
9.04
12.02
14.07
18.48 12
Researchers have conducted a survey of 1600 coffee drinkers asking how much coffee they drink in order to confirm previous studies. Previous studies have indicated that 72% of Americans drink coffee. At πΌ = 0.05, is there enough evidence to conclude that the distributions are the same?
GOODNESS-OF-FIT TEST
Reject π»π if Οπ is larger than the critical value (right-tailed test) Οπ (π β ππππ)
π. πππ
π β ππππ
>
π. ππ
There is enough statistical evidence to reject the null hypothesis and to believe that the old percentages no longer hold.
13
GOODNESS-OF-FIT TEST
2. A department store, A, has four competitors: B, C, D, and E. Store A hires a consultant to determine if the percentage of shoppers who prefer each of the five stores is the same. A survey of 1100 randomly shoppers prefer are as follows. Is there enough evidence using πΌ = 0.05 to conclude that the proportions are really the same?
Forbes.com 14
2. A department store, A, has four competitors: B, C, D, and E. Store A hires a consultant to determine if the percentage of shoppers who prefer each of the five stores is the same. A survey of 1100 randomly shoppers prefer are as follows. Is there enough evidence using πΌ = 0.05 to conclude that the proportions are really the same?
GOODNESS-OF-FIT TEST
Store Number of Shoppers
A 262
B 234
C 204
D 190
E 210
15
GOODNESS-OF-FIT TEST
2. A department store, A, has four competitors: B, C, D, and E. Store A hires a consultant to determine if the percentage of shoppers who prefer each of the five stores is the same. A survey of 1100 randomly shoppers prefer are as follows. Is there enough evidence using πΌ = 0.05 to conclude that the proportions are really the same?
π»π :The population frequencies are equal to the expected frequencies. π»π : The null is false. πΌ = 0.05 πππππππ ππ πππππππ: π β 1 = 5 β 1 = 4 16
2. A department store, A, has four competitors: B, C, D, and E. Store A hires a consultant to determine if the percentage of shoppers who prefer each of the five stores is the same. A survey of 1100 randomly shoppers prefer are as follows. Is there enough evidence using πΌ = 0.05 to conclude that the proportions are really the same?
GOODNESS-OF-FIT TEST
Preference
%of Shoppers
E
O
O-E
πΆβπ¬
A B C D E
20% 20% 20% 20% 20%
220 220 220 220 220
262 234 204 190 210
42 14 -16 -30 -10
1764 196 256 900 100
Οπ ΰ·
πΆβπ¬ π¬
π
πΆβπ¬ π¬
π
8.018 0.891 1.163 4.091 0.455
π
= ππ. πππ
17
Οπ = ππ. πππ
Percentage Points of the Chi-Square Distribution
0.50
0.25
0.10
0.05
0.01
1
0.455
1.32
2.71
3.84
6.63
2
1.386
2.77
4.61
5.99
9.21
3
2.366
4.11
6.25
7.81
11.34
4
3.357
5.39
7.78
9.49
13.28
5
4.351
6.63
9.24
11.07
15.09
6
5.348
7.84
10.64
12.59
16.81
7
6.346
9.04
12.02
14.07
18.48 18
GOODNESS-OF-FIT TEST
Reject π»π if Οπ is larger than the critical value (right-tailed test) Οπ (π β ππππ)
14.618 >
2. A department store, A, has four competitors: B, C, D, and E. Store A hires a consultant to determine if the percentage of shoppers who prefer each of the five stores is the same. A survey of 1100 randomly shoppers prefer are as follows. Is there enough evidence using πΌ = 0.05 to conclude that the proportions are really the same?
π β ππππ
9.49
There is enough statistical evidence to reject the null hypothesis and to believe that customers do not prefer each of the five stores equally.
19
TEST FOR INDEPENDENCE 20
TEST FOR INDEPENDENCE
β’ π»π : The two variables are independent β’ π»π : The two variables are not independent (dependent). β’ πΌ: level of significance β’ The expected frequency πΈπ,π =
ππ’π ππ π (π π’π ππ π) π πππππ π ππ§π
β’ πππππππ ππ πππππππ: (π β 1)(π β 1) β’
Οπ
=Ο
πΆβπ¬ π π¬
β’ From πΌ and π β 1, a critical value is determined from the chisquare table. β’ Reject π»π if Ο π is larger than the critical value (right-tailed test)
21
TEST FOR INDEPENDENCE
1. The result of a random sample of children with pain from musculoskeletal injuries treated with acetaminophen, ibuprofen, or codeine are shown in the table. At πΌ = 0.10 , is there enough evidence to conclude that the treatment and result are independent?
Dentagama.com 22
1. The result of a random sample of children with pain from musculoskeletal injuries treated with acetaminophen, ibuprofen, or codeine are shown in the table. At πΌ = 0.10 , is there enough evidence to conclude that the treatment and result are independent?
TEST FOR INDEPENDENCE
(r. 1) Significant Improvement
Acetaminophen (c. 1)
Ibuprofen (c. 2)
Codeine (c.3)
Total
58
81
61
200
66.7
66.7
(r. 2) Slight Improvement
42
19
33.3
33.3
Total
100
100
66.7
39
100
33.3
100
300 23
TEST FOR INDEPENDENCE
1. The result of a random sample of children with pain from musculoskeletal injuries treated with acetaminophen, ibuprofen, or codeine are shown in the table. At πΌ = 0.10 , is there enough evidence to conclude that the treatment and result are independent?
π»π :The population frequencies are equal to the expected frequencies. π»π : The null is false. πΌ = 0.05 πππππππ ππ πππππππ: 2 β 1 3 β 1 = 2 24
1. The result of a random sample of children with pain from musculoskeletal injuries treated with acetaminophen, ibuprofen, or codeine are shown in the table. At πΌ = 0.10 , is there enough evidence to conclude that the treatment and result are independent?
TEST FOR INDEPENDENCE
π
Row, Column
E
O
O-E
πΆβπ¬
1,1 1,2 1,3 2,1 2,2 2,3
66.7 66.7 66.7 33.3 33.3 33.3
58 81 61 42 19 39
-8.7 14.3 -5.7 8.7 -14.3 5.7
75.69 204.49 32.49 75.69 204.49 32.49
πΆβπ¬ π¬
1.135 3.067 0.487 2.271 6.135 0.975
π
Οπ ΰ·
πΆβπ¬
π
π¬
= ππ. ππ
25
Οπ = ππ. ππ
Percentage Points of the Chi-Square Distribution
0.50
0.25
0.10
0.05
0.01
1
0.455
1.32
2.71
3.84
6.63
2
1.386
2.77
4.61
5.99
9.21
3
2.366
4.11
6.25
7.81
11.34
4
3.357
5.39
7.78
9.49
13.28
5
4.351
6.63
9.24
11.07
15.09
6
5.348
7.84
10.64
12.59
16.81
7
6.346
9.04
12.02
14.07
18.48 26
TEST FOR INDEPENDENCE
1. The result of a random sample of children with pain from musculoskeletal injuries treated with acetaminophen, ibuprofen, or codeine are shown in the table. At πΌ = 0.10 , is there enough evidence to conclude that the treatment and result are independent?
Reject π»π if Οπ is larger than the critical value (right-tailed test) Οπ (π β ππππ)
14.07
π β ππππ
>
4.61
There is enough statistical evidence to reject the null hypothesis and to believe that there is a relationship between the treatment and the response.
27
TEST FOR INDEPENDENCE
2. The side effects of a new drug being tested against a placebo. A simple random sample of 565 patients yields the results as follows. At πΌ = 0.10 , is there enough evidence to conclude that the treatment is independent of the side effect of nausea?
Nutritionaloutlook.com 28
2. The side effects of a new drug being tested against a placebo. A simple random sample of 565 patients yields the results as follows. At πΌ = 0.10 , is there enough evidence to conclude that the treatment is independent of the side effect of nausea?
TEST FOR INDEPENDENCE
Result
Drug (c. 1)
Nausea (r. 1)
36
No nausea (r. 2)
254
Total
290
Placebo (c. 2)
Total
(25.15)
13
(23.85)
49
(264.85)
262
(251.15)
516
275
565
29
TEST FOR INDEPENDENCE
2. The side effects of a new drug being tested against a placebo. A simple random sample of 565 patients yields the results as follows. At πΌ = 0.10 , is there enough evidence to conclude that the treatment is independent of the side effect of nausea?
π»π : the treatment and response are independent. π»π : the treatment and response are dependent. πΌ = 0.10 πππππππ ππ πππππππ: 2 β 1 2 β 1 = 1 30
2. The side effects of a new drug being tested against a placebo. A simple random sample of 565 patients yields the results as follows. At πΌ = 0.10 , is there enough evidence to conclude that the treatment is independent of the side effect of nausea?
TEST FOR INDEPENDENCE
Row, Column
E
O
O-E
1,1 1,2 2,1 2,2
25.15 23.85 264.85 251.15
36 13 254 262
10.85 -10.85 -10.85 10.85
Οπ Ο
πΆβπ¬ π π¬
(πΆ
117.72 117.72 117.72 117.72
= ππ. ππ
πΆβπ¬ π¬
π
4.681 4.936 0.444 0.469
31
Οπ = ππ. ππ
Percentage Points of the Chi-Square Distribution
0.50
0.25
0.10
0.05
0.01
1
0.455
1.32
2.71
3.84
6.63
2
1.386
2.77
4.61
5.99
9.21
3
2.366
4.11
6.25
7.81
11.34
4
3.357
5.39
7.78
9.49
13.28
5
4.351
6.63
9.24
11.07
15.09
6
5.348
7.84
10.64
12.59
16.81
7
6.346
9.04
12.02
14.07
18.48 32
TEST FOR INDEPENDENCE
2. The side effects of a new drug being tested against a placebo. A simple random sample of 565 patients yields the results as follows. At πΌ = 0.10 , is there enough evidence to conclude that the treatment is independent of the side effect of nausea?
Reject π»π if Οπ is larger than the critical value (right-tailed test) Οπ (π β ππππ)
10.53
π β ππππ
>
2.71
There is enough statistical evidence to reject the null hypothesis and to believe that there is a relationship between the treatment and response.
33
TEST FOR HOMOGENEITY 34
TEST FOR HOMOGENEITY
β’ π»π : The distributions of the two populations are the same. β’ π»π : The distributions of the two populations are not the same. β’ πΌ: level of significance β’ The expected frequency πΈπ,π =
ππ’π ππ π (π π’π ππ π) π πππππ π ππ§π
β’ πππππππ ππ πππππππ: π β 1 β’
Οπ
=Ο
πΆβπ¬ π π¬
β’ From πΌ and π β 1, a critical value is determined from the chisquare table. β’ Reject π»π if Ο π is larger than the critical value (right-tailed test)
35
TEST FOR HOMOGENEITY 1. Do male and female college students have the same distribution of living arrangements? Use a level of significance of 0.05. Suppose that 250 randomly selected male college students and 300 randomly selected female college students were asked about their living arrangements: dormitory, apartment, with parents, other. The results are as follows. Do male and female college students have the same distribution of living arrangements?
Wcu.edu 36
1. Do male and female college students have the same distribution of living arrangements? Use a level of significance of 0.05. Suppose that 250 randomly selected male college students and 300 randomly selected female college students were asked about their living arrangements: dormitory, apartment, with parents, other. The results are as follows. Do male and female college students have the same distribution of living arrangements?
TEST FOR HOMOGENEITY
Dormitory (c. 1)
(r. 1) Males
72
Apartment (c. 2)
84
(r. 2) Females Total
91
49 (77.27)
(74.09)
86 (88.91)
163
With Parents (c. 3) (62.27)
88 (92.73)
170
Other (c. 4)
Total
45
250
(36.36)
35
(74.73)
(43.64)
137
80
300 550 37
TEST FOR HOMOGENEITY
π»π :The distribution of living arrangements for male college students is the same as the distribution of living arrangements for female college students. π»π : The distribution of living arrangements for male college students is not the same as the distribution of living arrangements for female college students. πΌ = 0.05 πππππππ ππ πππππππ: 4 β 1 = 3
1. Do male and female college students have the same distribution of living arrangements? Use a level of significance of 0.05. Suppose that 250 randomly selected male college students and 300 randomly selected female college students were asked about their living arrangements: dormitory, apartment, with parents, other. The results are as follows. Do male and female college students have the same distribution of living arrangements?
38
TEST FOR HOMOGENEITY
Row, Column
E
O
1,1 1,2 1,3
74.09 77.27 62.27
1,4 2,1 2,2 2,3 2,4
36.36 88.91 92.73 74.73 43.64
72 84 49 45 91 86 88 35
O-E
πΆβπ¬
π
4.372 -2.09 45.256 6.73 -13.27 176.165 74.587 8.64 4.372 2.09 45.256 -6.73 13.27 176.165 74.587 -8.64
πΆβπ¬ π¬
0.059 0.586 2.829 2.051 0.049 0.488 2.357 1.709
π
1. Do male and female college students have the same distribution of living arrangements? Use a level of significance of 0.05. Suppose that 250 randomly selected male college students and 300 randomly selected female college students were asked about their living arrangements: dormitory, apartment, with parents, other. The results are as follows. Do male and female college students have the same distribution of living arrangements?
Οπ ΰ·
πΆβπ¬
π
π¬
= ππ. πππ
39
Οπ = ππ. πππ
Percentage Points of the Chi-Square Distribution
0.50
0.25
0.10
0.05
0.01
1
0.455
1.32
2.71
3.84
6.63
2
1.386
2.77
4.61
5.99
9.21
3
2.366
4.11
6.25
7.81
11.34
4
3.357
5.39
7.78
9.49
13.28
5
4.351
6.63
9.24
11.07
15.09
6
5.348
7.84
10.64
12.59
16.81
7
6.346
9.04
12.02
14.07
18.48 40
Chi-Square Distribution Calculator
1. Do male and female college students have the same distribution of living arrangements? Use a level of significance of 0.05. Suppose that 250 randomly selected male college students and 300 randomly selected female college students were asked about their living arrangements: dormitory, apartment, with parents, other. The results are as follows. Do male and female college students have the same distribution of living arrangements?
41
TEST FOR HOMOGENEITY
Reject π»π if Οπ is larger than the critical value (right-tailed test)
πΌ = 0.05
1. Do male and female college students have the same distribution of living arrangements? Use a level of significance of 0.05. Suppose that 250 randomly selected male college students and 300 randomly selected female college students were asked about their living arrangements: dormitory, apartment, with parents, other. The results are as follows. Do male and female college students have the same distribution of living arrangements?
π· β πππππ
Type 0.05 > equation The distribution of living arrangements for male here. college students is not the same as the distribution of living arrangements for female college students.
42
TEST FOR HOMOGENEITY
2. In a study of the television viewing habits of children, a developmental psychologist selects a random sample of 300 first graders β 100 boys and 200 girls. Each child is asked which of the following TV programs they like best. The Long Ranger, Sesame Street, or The Simpsons. Results are shown in the contingency as follows.
punjabiCenter.com 43
2. In a study of the television viewing habits of children, a developmental psychologist selects a random sample of 300 first graders β 100 boys and 200 girls. Each child is asked which of the following TV programs they like best. The Long Ranger, Sesame Street, or The Simpsons. Results are shown in the contingency as follows.
TEST FOR HOMOGENEITY
Lone Ranger (c. 1)
(r. 1) Boys
50
(r. 2) Girls
50
Total
Sesame Street (c. 2)
30 (33.33)
20 (36.67)
80 (66.67)
100
The Simpsons (c. 3)
(73.33)
110
Total
100
(30)
70
200
(60)
90
300 44
TEST FOR HOMOGENEITY
2. In a study of the television viewing habits of children, a developmental psychologist selects a random sample of 300 first graders β 100 boys and 200 girls. Each child is asked which of the following TV programs they like best. The Long Ranger, Sesame Street, or The Simpsons. Results are shown in the contingency as follows.
π» of boys who prefer the Lone π»ππ ::The ππππ¦π proportion ππππ πΏπππ π
πππππ = π»πππππ ππππ πΏπππ π
πππππ Ranger is identical to the proportion of girls. π»π : ππππ¦π ππππ πππ πππ ππ‘ππππ‘ = π»πππππ ππππ πππ πππ ππ‘ππππ‘ Similarly, for the other programs. π»π : ππππ¦π ππππ π‘βπ πππππ πππ = π»πππππ πππππ‘βπ πππππ πππ π»π : At least one of the null hypothesis statement is false. πΌ = 0.05 πππππππ ππ πππππππ: 3 β 1 2 β 1 = 2
45
2. In a study of the television viewing habits of children, a developmental psychologist selects a random sample of 300 first graders β 100 boys and 200 girls. Each child is asked which of the following TV programs they like best. The Long Ranger, Sesame Street, or The Simpsons. Results are shown in the contingency as follows.
TEST FOR HOMOGENEITY
π
Row, Column
E
O
O-E
πΆβπ¬
1,1 1,2 1,3 2,1 2,2 2,3
33.33 36.67 30 66.67 73.33 60
72 30 20 50 80 70
16.67 -6.67 -10 -16.67 6.67 10
277.78 44.44 100 277.78 44.44 100
πΆβπ¬ π¬
8.33 1.21 3.33 4.17 0.61 1.67
π
Οπ ΰ·
πΆβπ¬
π
π¬
= ππ. πππ
46
Οπ = ππ. πππ
Percentage Points of the Chi-Square Distribution
0.50
0.25
0.10
0.05
0.01
1
0.455
1.32
2.71
3.84
6.63
2
1.386
2.77
4.61
5.99
9.21
3
2.366
4.11
6.25
7.81
11.34
4
3.357
5.39
7.78
9.49
13.28
5
4.351
6.63
9.24
11.07
15.09
6
5.348
7.84
10.64
12.59
16.81
7
6.346
9.04
12.02
14.07
18.48 47
Chi-Square Distribution Calculator
2. In a study of the television viewing habits of children, a developmental psychologist selects a random sample of 300 first graders β 100 boys and 200 girls. Each child is asked which of the following TV programs they like best. The Long Ranger, Sesame Street, or The Simpsons. Results are shown in the contingency as follows.
48
TEST FOR HOMOGENEITY
Reject π»π if Οπ is larger than the critical value (right-tailed test)
πΌ = 0.05
π· β πππππ
0.05
> 0.00005
There is enough statistical evidence to reject the null hypothesis and to believe that the proportion of boys who like Lone Ranger/Sesame Street/The Simpsons is not identical to girls who like the same TV programs, respectively.
2. In a study of the television viewing habits of children, a developmental psychologist selects a random sample of 300 first graders β 100 boys and 200 girls. Each child is asked which of the following TV programs they like best. The Long Ranger, Sesame Street, or The Simpsons. Results are shown in the contingency as follows.
49