The Pattern of Skewness And Kurtosis Using Mean Score And Logit In Measuring Adversity Quotient (AQ) For Normality Testing

Abstract

This study aims to examine the pattern of skewness and kurtosis using mean and logit in measuring AQ
for normality testing. Besides, the research also investigates the ability and potential of logit to meet the
acceptance range of normality. The study uses survey design with questionnaires. A total of 1,845
students from five Malaysian polytechnics were selected using a proportionate stratified multistage
cluster sampling. Instrument of IKBAR with 66 items of AQ and four CORE constructs were used after
evaluated by 37 experts and analysed by Rasch model. The findings show that of the logit of skewness and
kurtosis are higher than the mean. The difference in skewness shows the ranking of C-O-R-E while the
kurtosis is C-E-R-O. The skewness of mean and logit indicates that Ownership records the largest value.
The mean of kurtosis indicates Reach as the largest and Ownership for logit. Both skewness and kurtosis
show that Control as the smallest for mean and logit. The pattern of skewness and kurtosis for mean and
logit are in compliance with their best range. Results show that the skewness and kurtosis model have a
strong linear relationship between mean and logit. The models also explain more than 70% of the
variability of the response. Besides, the results mentioned about the linear moderate relationship between
the differences and rate of changes of skewness between mean and logit, but inversely for kurtosis. Based
on the results, it's clear that the patterns of logit have a great potential to be generating as normality
testing besides mean. Nevertheless, the logit is greater than the mean and will be more difficult to meet
the requirement of normality acceptance range. Further studies are proposed to explore the potential of
logit from different contexts of study as well as suggesting the acceptance of specific normality ranges for
logit.