Written Corrective Feedback (WCF) is used extensively in second-language (L2) writing classrooms despite controversy over its effectiveness. This study examines indirect WCF, an instructional procedure that flags L2 students' errors with editing symbols that guide their corrections. WCF practitioners assume that this guidance will lead to increased grammatical competence over time in new writing samples. This study finds that these assumptions are correct overall. However, in-depth analyses of L2-English learners' correction behaviors in four elicitation tasks over a 12-week period demonstrate that WCF is not uniformly effective at increasing accuracy for all grammatical constructions. In fact, WCF fails to exert any positive effect with a number of grammatical constructions. This result can be understood via Skill Acquisition Theory (SAT) when the treatability of constructions with WCF is considered. Specifically, grammatical constructions that include only a binary option for correct usage are highly amenable to positive change via WCF since employing WCF is akin to correcting errors flagged on a true/false test. By contrast, grammatical constructions with more than a binary choice for correct usage, akin to correcting errors flagged on a multiple-choice test, are not amenable to positive change.