PERBANDINGAN TEKNIK BERGRAF UNTUK DATA BERKATEGORI

dc.contributor.author	PING, HOO LING
dc.date.accessioned	2016-02-03T06:13:28Z
dc.date.available	2016-02-03T06:13:28Z
dc.date.issued	2002-02
dc.description.abstract	Graphical methods for categorical data are still in the period of infancy. Not many graphical methods have been introduced. Besides, methods which are discussed in the literature are not available in nowadays statistics software. Thus, graphical methods for categorical data are not used widely. Therefore, this dissertation aims to introduce the concept for two types of graphical methods, which are suitable for the interpretation of categorical data. The two types of graphical methods are correspondence analysis and mosaic display. Next, the results from the two methods will be compared. Correspondence analysis analyzes the relationship between not only two categories of discrete variable but also for several categories at the same time. It displays the results visually as points within a two-dimensional space that facilitates interpretation. Mosaic display is a graphical method for n-way contingency tables. It presents each cell directly or indirectly with tiles and the areas are proportional to the cell frequencies. For this dissertation, an analysis of a survey using videotape as an alternative to present the courses to the students of The School of Distance Education, USM has been carried out. By using the methods of correspondence analysis and mosaic display, we have found that in video 1 and video 2, too many difficult concepts are introduced. Video 3, video 4, video 5 and video 6 are suitable for presentation. From the research that have been done, the. methods of correspondence analysis and mosaic display can be used to present the relationships of variables from categorical data during the initial stage before further study is done.	en_US
dc.identifier.uri	http://hdl.handle.net/123456789/1752
dc.subject	PERBANDINGAN TEKNIK BERGRAF UNTUK DATA BERKATEGORI	en_US
dc.title	PERBANDINGAN TEKNIK BERGRAF UNTUK DATA BERKATEGORI	en_US
dc.type	Thesis	en_US