The purposes of this research were to: 1) to compare the effects of different sequences of feedback types on the students' writing ability and their effect size; 2) to compare the effects of the levels of general English proficiency (high, moderate, and low) on the students' writing ability and their effect size; 3) to investigate the interaction effects between the different sequences of feedback types and levels of general English proficiency on the students' writing ability; and 4) to survey the students preferences for self correction, paper-pencil peer feedback, and the integration of Facebook and the peer feedback strategy. This study was of an experimental, 3x3 factorial design and was conducted with 108 engineering students with different levels of general English proficiency. The students were trained in how to do self-correction or how to provide effective peer feedback before doing the assigned tasks. Two-way ANOVA was employed for the data analysis for the first three research questions, and descriptive statistics and content analysis were used for analyzing the data from the questionnaires, interviews, and reflective journals for the fourth research question. Pedagogical implications for EFL writing classes are discussed.