Abstract

MENDOZA MENDEZ, Rafael V.; DORANTES CORONADO, Ernesto Joel; CEDILLO MONROY, José  and  JASSO ARRIAGA, Xóchitl. Discriminant Function Analysis as interpretive tool for the Study on Mobile Phone Overuse in students of the Bachelor's degree in Administrative Data Processing of the Autonomous University of Mexico State, University Center UAEM Temascaltepec. RIDE. Rev. Iberoam. Investig. Desarro. Educ [online]. 2017, vol.7, n.14, pp.222-247. ISSN 2007-7467.  https://doi.org/10.23913/ride.v7i14.282.

The aim of this article was to demonstrate the effectiveness of the Multiple Discriminant Analysis (MDA) statistical method, as a technique for the interpretation of the results of a study conducted in the Autonomous University of Mexico State, University Center UAEM Temascaltepec in students of the Bachelor's degree in Administrative Data Processing about Mobile Phone Overuse, also known as Smartphone Addiction. The levels of addiction were gender-differentiated with precision, which allowed classifying new cases, and due to their characteristics make the most of the relations between the independent variables to differentiate them more efficiently. “The Discriminant Analysis starts from a Data Table of "n" individuals, who have been measured by "p" independent quantitative variables, that act as the characteristics of each profile” (Carvajal, Trejos, and Soto Mejía, 2004).

With this method the Discriminant Analysis provided that only 100 cases out of 112 students surveyed were valid, corresponding to the 89.3% of the total sample; also, the group statistics reported 66 cases male and 34 female, and the statistic of the contrast of global significance Wilks's Lambda was moderately high (0.621). Table analysis of variance with statistical F allowed contrasting the hypothesis of equality of means between groups, sensing that only two variables make the difference in the representation of gender: Use of mobile devices (P≤0. 042) and frequency of use of the mobile device (P≤0. 036). The logarithmic sample of the determinants of all matrices used in the statistical calculation showed overlap occurs in both groups, which indicates that they tend to equality: Do you use a mobile device?, women gained 59.46% compared to 36.49% of men, while in the item: How often do you use your mobile device?, women reached 72.97% in contrast to 54.05% of male. Finally, the Multiple Discriminant Analysis method showed its effectiveness to find differences in the addiction by gender.

Keywords : statistical methods; Multiple Discriminant Analysis (MDA); Mobile phone overuse; Temascaltepec; Administrative Data Processing (ADP).