an easy task….
Pearson chi square test also called as chi square test for independence or chi square test of
association is used to find whether there is a relationship between two categorical variables.
To carry out chi square test, null hypothesis is considered first. In SPSS, when cross
tabulation is done and two variables are considered, null hypothesis is made. Null hypothesis
states that no association exists between the two cross tabulated values considered and thus
the variables are statistically independent. For example, if 2 methods are compared to find out
which one is better and statement is made that both methods are equally good then this is
called as null hypothesis. If it is proposed that two variables are related then it is considered
as alternative hypothesis. For example, if one method is superior than the other, then called as
alternative hypothesis. Through cross tabulation and statistics involved in it, chi square value
table including significant difference value is calculated and studied to find the acceptance or
rejection of null hypothesis. If significant difference is less than 0.05 (generally taken
between 0-1) then null hypothesis is rejected which means there is a relationship between the
two variables considered. Significant difference is generally looked at 0.05 which denotes
that the confidence level is 95%, it can be taken at any value between 0-1. If significant
difference is more than 0.05, then null hypothesis is accepted, which signifies that there is no
relationship between the two variables considered.
Example : Null hypothesis – Area of residence and BMI (Body mass index) are
independent.
Alternative hypothesis- Area of residence and BMI are dependent.
Or Null hypothesis- There is no association between living in an urban area and being
overweight.
i.e. It is not that people living in urban area are overweight than people living in rural area.
Here, significance level is 0.05 is taken. Cross tabulation using SPSS is done, in which
people residence is taken in rows and weight category is taken in columns. Statistics is
applied and chi square table including significant difference is studied. In this example,
significant value in pearson chi square is obtained as 0.0001 which is much less than 0.05
and thus signifies that null hypothesis is rejected i.e. two variables considered (weight and
residential area ) are strongly related than people living in urban area are overweight than
people in rural area.
By
Shailja Talwar
