A Linear Regression Model for Assessing the Ranking of Web Sites Based on Number of Visits

Many web sites are designed and established without sufficient professional skills and resources. The quality of these websites is often dubious. Therefore, how to evaluate the quality of a web site has become an important issue. In this work, the stepwise regression method is applied to assess the ranking of web sites for two different categories. The ranking is based on the average number of visits per day. A total of fourteen factors frequently found in the literature are considered as independent variables in developing the model. The regression analysis result shows that the regression models differ for two different categories of web sites, but there are three variables common to the two resulting models, Latest update, Broken links, and Help features. The average prediction accuracies of both models exceed 75%.