TEST OF MOBILE PHONES SCREEN DATA BASED ON THE MAHALANOBIS DISTANCE

Abstract

A survey was conducted to obtain data set on scratched mobile smart and ordinary (conventional) phones screen, and to further investigate if the data set obtained are normally distributed or not. The Mahalanobis distance, the correlation coefficient, Chi squared and the quantile quantile plots are applied to determine if the data set obtained based on the categorization is normally distributed or otherwise. The techniques revealed that the data set are not normally distributed. The mean and the standard deviation approach based on the concept of data contamination validated the previous conclusions. At  level of significance, the hypothesis was rejected, implying non-normality of the data set. In general, the conclusions based on all techniques indicated that the data set are not normally distributed. This implies that though users of mobile phones are both diligent and non-diligent alike