This paper investigates undergraduate students’ revision processes and their impacts on text quality by using mixed-methods. For this purpose, 50 undergraduate students (48 available among them) participated and they offered written outputs and video capture files of their own writing process during computer writing. Analysis was conducted in terms of the transformative design; that is, quantification of the qualitative data. Data coding was done in two stages; first, the qualitative coding of the revising process, then the quantification based on the fist-stage coding and memos written by coders. Quantification was conducted by 5-point scales for 4 components: editing (editing during the first draft), reading (reading text produced so far during the first draft), proofreading (editing and reading during the final draft), and effectiveness (the amount of textual change). Based on this quantification, correlation analysis, one-way ANOVA, EFA, and cluster analysis were conducted to verify the relations between variables, differences in text quality between levels of activities, and actual categories of revising activities. As a result, no meaningful difference in text quality was found between the levels of revising activities, and correlations between activities were relatively low or not high. Therefore, it can be concluded that behavioral aspects of revision (e. g. amount or pattern) do not affect text quality; we should find another factor for effective revising strategies, considering several studies (Galbraith & Torrance, 2004; Kieft, et al., 2007; Torrance, et al., 1999, 2000) insisted revising should be one of the two main writing strategies, along with planning.