Recognizing the importance of learning reflections, scholars and researchers classified reflections into various levels and types. Due to the diversity of reflective practice, learners are expected to perform differently, leading to a variety of evaluation methods and ways of classification. In the context of Web-based portfolio assessment, however, what levels and types of reflection take place, and in which way it can be appropriately sorted out? Besides, another issue remained unanswered is concerned with the representations of reflective behaviors and its applicable methods of evaluation. Rama and Battistoni (2001) suggested that the number and duration of reflective practice can serve as powerful indicators in portfolio learning. Thus, to evaluate reflections assessors should not only take into account the quality of reflection work, but also word count of reflective journals as well as the number and duration of peer reviews. On the other hand, a study conducted by Irby and Brown (1999) revealed that students shared a similar length of reflective journals, drawing our attention to the question about whether or not there are patent differences between individual learner in reflective behaviors.