Data Analysis of Senior Computer Literacy

Abstract

To outline the situation in area of seniors computer literacy in given sample of the population was the objective of this paper. More precisely, it deals with the level of a computer literacy of seniors and with classification this level on the basis monitored attributes. The data collection was realised on the basis cooperation with the senior centre and senior's clubs in city Pardubice. In total 407 seniors participated in this data survey. The data was evaluated in MS Excel and new output attribute based on the European Computer Driving Licence concept was created. The attribute represent a preparation of respondents to obtain the European Computer Driving Licence Certificate on the basis of computer skills. Seniors that do not use computer or notebook were discarded from the data set. The final data set included 252 seniors described by 29 attributes. Classification models were created in IBM SPSS Modeler on the basis new output attribute and selected inputs e.g. sociodemographic attributes, barriers in use of the Internet and purpose of computer use. Algorithms of decision trees as C5.0, CRT, CHAID, QUEST were chosen for solution of mentioned task. The best results of classification were achieved by C5.0 with boosting. The accuracy rate on the testing data set was 91.74%.