Computer evaluation in statistics for undergraduates

Abstract

The advance of the use of computers in the students' evaluation, mainly because of the high growth of distance education and large-scale evaluation systems, means that there is a very promising environment to develop new ideas in computer software with educational approaches (Bennett, 2006). This kind of evaluation system can be used both on high-stake tests as in low-stake tests, as well for self evaluation and better guides into the self adaptive learning. Not just that, this kind of evaluation strengthens the adaptive test development, increasingly frequent nowadays (Gershon, 2005). Although the implementation of effective systems requires a very adequate computational environment, that does not just transposes a paper test to the computer screen, but improves the way the students assimilate the test information and develop their learning while answering the test (Frye & Soloway, 1987). In this project, we investigate the best ways we can apply a computer system evaluation into the statistics teaching in our university, using contents such as interactive media and different types of user interaction with the items. We are also trying to give to the educator the possibility of building not just classical tests, using known correction models as item sum and percent of correction, but adaptive models too, like using the Item Response Theory (Weiss & Yoes, 1991) . This is going to produce an open software, made to be used with an internet browser, which facilitates its portability into different platforms. The more expansible as possible to help the community to its further development and usage in other environments next to other realities, such as high schools, different universities or within arbitrary subjects, as well as to work as a framework to build and test new kinds of evaluation models.