Question

observed	responsible	not responsible
illiterate	21	45
read write	43	95
primary	231	529
secondary	90	228
university	78	166

the following obsereved data from sample of drivers the question that can we test the kai square as 5*2 matrix or test each column seperatly and find the significancy of the relation between level of education and the responsibe and non responsible column of accidents.   

Answer 1

We can test the Chi-square as 5*2 matirx.(Not necessary to test each column seperately)

Chi-square test for association (or independence) can be to discover if there is relationship between two categorical variables.

For the given problem,

Categorical variables are :

Variable 1 : Level of education

Variable 2 : Accident responsibility (responsibe and not responsible of accidents)

The null hypothesis would be : H_o : Level of education and Responsible for accident are not related (Independent)

Alternate Hypothesis : H_a : Level of education and Responsible for accident are related (Dependent)

The test statistic :

O: Observed Data

E : Expected Data

Expected data :