Engineering educators and industry professionals recognize the need for graduates to be effective communicators, yet the effective teaching of communication remains a persistent contemporary issue, with studies continuing to indicate that engineering graduates are insufficiently prepared for workplace communication. Despite compelling arguments that that writing should be treated as a situated activity, writing instruction is often separated from content instruction within engineering curricula. Even when they are integrated, it is often in a way does not optimally support improvement of students' writing skills. Writing studies scholarship identifies best practices that include treating writing as a process, with pedagogy that includes writing and revising drafts based on feedback and revision. However, writing assignments in engineering courses often not process-oriented.

Challenges in addressing this problem include epistemology (i.e. instructors believe that learning to write and learning to engineer are separate skills), self-efficacy (i.e. instructors not feeling qualified to effectively provide feedback or writing instruction), and resources (i.e. inclusion of feedback and revision is unfeasible within key constraints of many engineering courses – limited instructor time and large student-faculty ratios).

A potential solution is to use peer feedback, where students provide each other feedback on drafts, which can support a process approach while addressing these challenges. Research outside engineering has demonstrated that peer feedback can be as or more effective than instructor feedback; to bring that work into engineering, this study examines peer feedback in the context of a first-year engineering course through a quasi-experimental intervention in which feedback and revision were incorporated into an existing assignment using several variations of peer feedback. Interventions included two types of feedback training as well as feedback from single peers and multiple peers.

Results support peer feedback in this context: it was statistically indistinguishable from instructor feedback when students were given sufficient instruction. Feedback from multiple peers, in fact was more effective than instructor feedback in improving writing quality, and in-class instruction was more effective than a handout only in helping students provide effective feedback. However, some general feedback recommendations – notably that readerly feedback should be encouraged directive feedback discouraged – were not supported. While writing studies literature encourages feedback that takes the position of the reader, readerly peer feedback reduced revision quality in this study. Directive feedback, on the other hand, caused improvements in writing quality, supporting previous hypotheses that the tightly-constrained genres in which engineers write justify more use of directive feedback.