Relationship between academic performance and usage pattern of non-academic websites by gender group

There is growing perception in academic community that internet and web resources are dominantly used for non-academic and social purposes. There have been some studies to validate this perception. Major objective of our analysis was to investigate the relationship between usage patterns of web sites which do not contribute directly to curricular activities and their relationships with academic performances of students from gender groups This paper presents results of an analysis of log files of a technical institute's proxy server for a period of over one year. The contents of log files were pre-processed to hide the identities of users and to create relational database of fields of our interests. We have implemented algorithms for statistical analysis and validated our results using Rapid-miner. Our results confirm the hypothesis that the students dominantly use internet resources for activities which do not directly contribute to curricular program. However, a very significant percentage of this non-academic usage is directed to desirable extracurricular activities such as news, sports, media and online games etc. It is interesting to note that the female students dominantly use social networking sites where as male students dominantly use entertainment websites. A very small percentage (9%) of the total time is spent on socially unacceptable and undesirable websites which include drugs and adult websites. Further, the academic performance, measured in terms of average CPI (Cumulative Performance Index), of all internet users is 7.5 and this group of students spent fifty five minutes on an average daily on the internet. In contrast, the average time spent in non-academic websites are 90 minutes daily on average. Average CPI of students using academic and nonacademic websites are 7.6 and 7.8 respectively which are almost same as that of all internet users. These average CPI numbers suggest that composition of student groups in terms of CPI for academic and non-academic websites are more or less similar.

Keywords: Social networks, Entertainment websites, Academic Performance, Log File