Page 76 - IJET_July2021_final
P. 76
From table-6 it is evident that the Urban) should be independent of
statistical value of Kolmogorov-Smirnov each other i.e. selection of subjects of
test of the Readiness Scores of rural one sample should not be affected by
participants and urban participants the selection of subjects of another
towards blended learning approach sample. It is clear from the observation
are 0.127, 0.123 respectively, whose of independent variable, there is
probability of significance at df (224) independence of the observation in
, df (213) is 0.000 which is less than collecting rural, urban and semi urban
0.01, hence significant at 0.01 level of groups.
significance. In this perspective the null
hypothesis “The given distribution of Assumption#2 The scores of
the readiness scores of rural and urban dependent variables must be at least
participants towards a blended learning on ordinal scale
approach do not differ significantly from On the basis of observation of the
the normal distribution” is rejected. scores of dependent variables, it is clear
Therefore, it can be concluded that the that these scores are on interval scale.
assumption of normality of readiness Therefore, the second assumption of
scores of rural and urban participants Kruskal Wallis H Test that the scores of
towards a blended learning approach is dependent variables must be at least on
not satisfied or it is violated.
ordinal scale, is also fulfilled.
Further, the above description
indicates that the readiness scores of Assumption#3 Uniformity of
rural and urban participants are not distribution of dependent variable at
normally distributed and readiness both levels of independent variable
scores of semi urban participants are This assumption holds whether the
normally distributed. Because the first distribution of scores of dependent
assumption of the normality of data variables for all the groups of
of one-way ANOVA is not satisfied at independent variables (Rural, Urban
all the levels of dependent variables, and Semi Urban) has the same shape
hence we use Kruskal Wallis H Test for or a different shape. For testing it,
further analysis which is an alternative graphical representation was used and
non parametric statistics technique uniformity found.
of one- way ANOVA. Further, before
applying Kruskal Wallis H Test, all the Assumption#4 Equality of variances
assumptions related to Kruskal Wallis (Non parametric)
H Test were also tested which are given
below: For testing this assumption non
parametric levene’s test for equal
Assumption#1- Independence of the variance is used. The results of ANOVA
Groups related to this assumption are presented
in the following table -7.
According to this assumption all the
three samples (Rural, Urban and Semi
66 Indian Journal of Educational Technology
Volume 3, Issue 2, July 2021