To test whether the two independent estimates of population variances are significantly different or not, or are uniform or not to test whether the two independent estimates of the samples have come from the same universe and have a common variance, we use variance ratio set. Suppose that n_{1},n_{2} are samples drawn from a normal population with corresponding means µ_{1}, µ_{2} and variance (σ1^{2},) (σ2^{2}). Then the following steps are followed to test the variance ratio test.
Step I: Null hypothesis. H0:σ_{1}^{2},=σ_{2}^{2}
That is no significant difference between two population variance or there is no significant difference between two the independent estimates of the common population variance or two population variance are same.
Step 2. Alternative hypothesis.H_{1}≠σ_{2}^{2}=σ^{2}
That is there is significance difference between two population variances, or there is the significant difference between two independent estimates of the common population variance, or two population variance are not same.
Step:3 Test statistic
Under H_{0}, the statistic is.
$$F=F=\frac{S_1^2}{S_2^2}∼F(V_1.V_2)IfS_1^2>S_2^2$$
$$and\,F=\frac{S_2^2}{S_1^2}∼F(V_2.V_1)IfS_2^2>S_1^2$$
$$Where\,S_1^2=Unbiased\,estimate\,of\,first\,population\,Varience$$
$$Where\,S_2^2=Unbiased\,estimate\,of\,second\,population\,Varience$$
If raw data is given, in this situation, we use following formula to calculate unbiased estimate of population variences.
$$S_1^2=\frac{1}{n_1-1}\sum(X_1-\overline{X}_1)^2[Actual\,mean\,Method]$$
$$=\frac{1}{n_1-1}[\sum(X_1^2-\frac{\sum(X_1)^2}{n_1}]\,[Direct\,method]$$
$$=\frac{1}{n_1-1}[\sum(d_1^2-\frac{\sum(d_1)^2}{n_1}]\,[Shortcut\,method]$$
$$S_2^2=\frac{1}{n_2-1}\sum(X_2-\overline{X}_2)^2[Actual\,mean\,Method]$$
$$=\frac{1}{n_2-1}[\sum(X_2^2-\frac{\sum(X_2)^2}{n_2}]\,[Direct\,method]$$
$$=\frac{1}{n_2-1}[\sum(d_2^2-\frac{\sum(d_2)^2}{n_2}]\,[Shortcut\,method]$$
$$Where\,d_1=X_1-A_1\,,A_1=assumed\,mean\,taken\,from\,X_1$$
$$Where\,d_2=X_2-A_1\,,A_2=assumed\,mean\,taken\,from\,X_2$$
If biased results I,e s_{1} and s_{2} or (s12 and s22) are given, in this situation, we use the following the formula to calculate unbiased estimates of population variance
$$S_1^2=\frac{n_1S_1^2}{(n_1-1)}$$
$$S_2^2=\frac{n_2S_2^2}{(n_2-1)}$$
$$Here\,,S_1^2=sample\,varience\,of\,first\,population\,variable$$
$$S_1^2=\frac{1}{n_1}\sum(X_1-\overline{X}_1)^2[Actual\,mean\,Method]$$
$$=\frac{1}{n_1}[\sum(X_1^2-\frac{\sum(X_1)^2}{n_1}]\,[Direct\,method]$$
$$=[\sum(d_1^2-\frac{\sum(d_1)^2}{n_1}]\,[Shortcut\,method]$$
$$S_2^2=\frac{1}{n_2}\sum(X_2-\overline{X}_2)^2[Actual\,mean\,Method]$$
$$=\frac{1}{n_2}[\sum(X_2^2-\frac{\sum(X_2)^2}{n_2}]\,[Direct\,method]$$
$$=\frac{1}{n_2}[\sum(d_2^2-\frac{\sum(d_2)^2}{n_2}]\,[Shortcut\,method]$$
Step:4 Level of significance.
We have to set 1% or 5% level os significance.(α)
Step:5 Degree of freedom:
$$(n_1-1,n_2-1)\,If\,F=\frac{S_1^2}{S_2^2}$$
$$or\,n_2-1,n_1-1),If\,F=\frac{S_1^2}{S_2^2}$$
Step 6. Critical value.
We have to determine the tabulated value of F at (α)% level of significance for required degree of freedom from F table.
Step 7. Decision.
If the calculated value of F(≤) tabulated value of F, we accept null hypothesis otherwise, we reject null hypothesis I,e accept the alternative hypothesis.
Kerlinger, F.N. Foundation of Behavioural Research. New Delhi: Surjeet Publication, 2000.
Kothari, C.R. Research Methodology. India: Vishwa Prakashan, 1990.
Singh, M.L. and J.M Singh. Understanding Research Methodology. 1998.
Singh, Mrigendra Lal. Understanding Research Methodology. Nepal: National Book centre, 2013.
$$(n_1-1,n_2-1)\,If\,F=\frac{S_1^2}{S_2^2}$$
$$or\,n_2-1,n_1-1),If\,F=\frac{S_1^2}{S_2^2}$$
If the calculated value of F(≤) tabulated value of F, we accept null hypothesis otherwise, we reject null hypothesis I,e accept the alternative hypothesis.
No discussion on this note yet. Be first to comment on this note