IMPLEMENTASI IMRPOVED CHI-SQUARE AUTOMATIC INTERACTION DETECTION PADA KLASIFIKASI FAKTOR-FAKTOR YANG MEMPENGARUHI LITERASI INFORMASI GENERASI MUDA

Zian  Bula; Resmawan Resmawan; La Ode  Nashar; Salmun K.  Nasib

doi:10.21009/JSA.06207

Zian Bula Prodi Statistika, Universitas Negeri Gorontalo, Bone Bolango 96119, Indonesia
Resmawan Resmawan Department of Mathematics, Universitas Negeri Gorontalo
La Ode Nashar Prodi Statistika, Universitas Negeri Gorontalo, Bone Bolango 96119, Indonesia
Salmun K. Nasib Prodi Statistika, Universitas Negeri Gorontalo, Bone Bolango 96119, Indonesia

Keywords: Classification, Improved CHAID, Classification Performance, Information Literacy

Abstract

Information Literacy skills are needed to find quality sources and manage and sort information so that it can be used to improve the quality of life and community empowerment. The number of factors that affect information literacy causes the need for classification. The method used is Improved Chi-Square Automatic Interaction Detection (Improved CHAID), which aims to classify influencing factors with Information Literacy abilities. This study uses primary data, namely 237 Mananggu Young Generation (15-24 years), with Information Literacy as the dependent variable. The independent variables consist of reading interest, reading habits, gender, digital literacy, information needs, critical thinking, and information-seeking behavior. Based on the Improved CHAID analysis, the factors that significantly affect information literacy are Reading Habits (83%), Information Needs (89%), and Critical Thinking (94%). The classification performance of Testing Data is 40%, with a classification accuracy of 77% or from 95 samples, there are 73 samples that are properly classified. The sensitivity of 78% shows that the classification results are able to predict samples that have information literacy, 74% specificity indicates that the classification results are able to predict samples that do not have information literacy, and Press's Q 27.38 indicates a stable classification or statistically significant.