Exercise.

Chi square test – test for association and goodness of fit

– test for goodness of fit

If 0i, (i=1,2,…..,n) is a set of observed (experimental frequencies )and Ei (i=1,2,…..,n) is

~ df

Example1:

The number of yiest cells counted in a haemocytometer is compared to the theoretical value is
given below. Does the experimental result support the theory?

No. of Yeast cells Obseved Frequency Expected Frequency

0 103 106

1 143 141

2 98 93

3 42 41

4 8 14

5 6 5

Solution:

H0: the experimental results support the theory

H1: the esperimental results does not support the theory.

Level of significance=5%

Test Statistic:

1
 ~ df

Oi Ei Oi-Ei (Oi-Ei)2 (Oi-Ei)2/Ei

103 106 -3 9 0.0849

143 141 2 4 0.0284

98 93 5 25 0.2688

42 41 1 1 0.0244

8 14 -6 36 2.5714

6 5 1 1 0.2000

400 400 3.1779

∴ =3.1779

Table value:

(6-1=5 at 5 % l.os)= 11.070

Inference

cal< tab

We accept the null hypothesis.

test for independence of attributes

Formula:

= ~ df

Oij – observed frequencies

Eij – Expected frequencies

2
 m= number of rows

n= number of columns

Exapmle 2:

The severity of a disease and blood group were studied in a research projest. The findings
sre given in the following table, knowmn as the m xn contingency table. Can this severity of the
condition and blood group are associated.

Severity of a disease classified by blood group in 1500 patients.

Blood Groups
Condition Total
O A B AB

Severe 51 40 10 9 110

Moderate 105 103 25 17 250

Mild 384 527 125 104 1140

Total 540 670 160 130 1500

Solution:

H0: The two attributes severity of the condition and blood groups are not associated.
H1: The two attributes severity of the condition and blood groups are associated.

Calculation of Expected frequencies

Blood Groups
Condition Total
O A B AB

Severe 39.6 49.1 11.7 9.5 110

Moderate 90.0 111.7 26.7 21.7 250

Mild 410.4 509.2 121.6 98.8 1140

3
 Total 540 670 160 130 1500

Test statistic:

= ~ df

Here m= 3.,n =4

Calculations:

Oi Ei Oi-Ei (Oi-Ei)2 (Oi-Ei)2/Ei

51 39.6 11.4 129.96 3.2818

40 49.1 -9.1 82.81 1.6866

10 11.7 -1.7 2.89 0.2470

9 9.5 -0.5 0.25 0.0263

105 90.0 15 225.00 2.5000

103 111.7 -8.7 75.69 0.6776

25 26.7 -1.7 2.89 0.1082

17 21.7 -4.7 22.09 1.0180

384 410.4 -26.4 696.96 1.6982

527 509.2 17.8 316.84 0.6222

125 121.6 3.4 11.56 0.0951

104 98.8 5.2 27.04 0.2737

12.2347

∴ =12.2347

Table value:

(3-1)(4-1)= (6) at 5 % l.os)= 12.59

4
Inference

cal< tab

We accept the null hypothesis.

= ~ (2-1) (2-1) df = (1) df

Example 3:

In order to determine the possible effectof a chemical treatment on the rate of germination of
cotton seeds a pot culture experiment was conducted. The results are given below

Chemical treatment and germination of cotton seeds

Germinated Not germinated Total

Chemically Treated 118 22 140

Untreated 120 40 160

Total 238 62 300

Does the chemical treatrment improve the germination rate of cotton seeds at 1 % level?

Solution:

H0:The chemical treatment does not improve the germination rate of cotton seeds.
H1: The chemical treatment improves the germination rate of cotton seeds.

L.O.S = 1 %

Test statistic

= ~ (2-1) (2-1) df = (1) df

5
 =

Table value:

(1) df at 1 % L.O.S = 6.635

Inference

cal< tab

We accept the null hypothesis.

(i.e)The chemical treatmentwill not improve the germination rate of cotton seeds significantly.

Yates correction for continuity

Then use, = ~ (1) df

(or)

Directly use the - statistic as

= ~ (1) df

Example 4:

In an experiment on the effect of a growth regulator on fruit setting in muskmelon the

Fruit set Fruit not set Total

Treated 16 9 25

6
 Control 4 21 25

Total 20 30 50

Solution:

H0:Fruit setting in muskmelon does not depend on the application of growth regulator.

H1: Fruit setting in muskmelon depend on the application of growth regulator.

L.O.S = 1 %

Tet statistic

= ~ (1) df

Table value:

(1) df at 1 % L.O.S = 6.635

Inference

cal > tab

We reject the null hypothesis.

(i.e) Fruit setting in muskmelon is influenced by the growth regulator.

Learning Exercise

1. The theory predicts the proportion of beans in the 4 groups A, B, C, D should be 9:3:3:1. In an
experiment among 1600 beans, the number in the four groups were 882, 313, 287 and 118. Does
the experimental result support the theory.

7
2. A study was conducted, among 100 professors from 3 different divisions for the preference on
beverages of 3 categories test if there is any relationship between the field of teaching and
preference of beverage.

Field of teaching

Beverage Business Social Agri Total

Tea 20 10 10 40

Coffee 10 10 15 35

Cold 10 8 7 25
drinks

Total 40 28 32 100

3. A random sample of 600 students from Delhi University are selected and asked their opinion ab
out autonomous Status of Colleges. The results were given below. Test the hypothesis at 5%
level that opinions are independent of class groupings.

Class Favour of Against

Commerce 120 80 200

Science 130 70 200

Arts 70 30 100

Total 400 200 600

4. In a survey of preference of new coverage 100 pearsons are collected and taste preference of
average was surveyed according to sex of the person. We conclude that the taste preference and
sex of the person are associated.

Male Female

8
 Favour 35 25 60

Against 25 15 40

Total 60 40 100

5. Two new food stuffs were introduced and public opinion was sought based on the taste of the food
stuff. The results are given below. Examine whether there is an association between the category
of the food stuffs and the taste of the food stuffs.

Category

A B

Tasty 620 380 1000

Not tasty 550 450 1000

Total 1170 830 2000