Text-driven, quantitative methods provide new ways to analyze student writing, by uncovering recurring grammatical features and related stylistic effects that remain tacit to students and those who read and evaluate student writing. To date, however, these methods are rarely used in research on students transitioning into US postsecondary writing, and especially rare are studies of student writing that is already scored according to high-stakes writing expectations. This study offers a corpus-based, comparative analysis of higher- and lower-scoring Advanced Placement (AP) exams in English, revealing statistically significant syntactic patterns that distinguish higher-scoring exams according to “informational production” and lower-scoring essays according to “involved” or “interactional” production (Biber, 1988). These differences contribute to what we label emphatic generality in the lower-scoring essays, in which writers tend to foreground human actors, including themselves. In contrast, patterns in higher-scoring essays achieve what we call elaborated specificity, by focusing on and explicating specific, often abstract, concepts.These findings help uncover what is rewarded (or not) in high-stakes writing assessments and show that some students struggle with register awareness. A related implication, then, is the importance of teaching register awareness to students at the late secondary and early university level—students who are still relative novices, but are being invited to compose informationally dense prose. Such register considerations, and specific features revealed in this study, provide ways to help demystify privileged writing forms for students, particularly students for whom academic writing may seem distant from their own communicative practices and ambitions.

Text-driven, quantitative methods provide new ways to analyze student writing, by uncovering recurring grammatical features and related stylistic effects that remain tacit to students and those who read and evaluate student writing. To date, however, these methods are rarely used in research on students transitioning into US postsecondary writing, and especially rare are studies of student writing that is already scored according to high-stakes writing expectations. This study offers a corpus-based, comparative analysis of higher- and lower-scoring Advanced Placement (AP) exams in English, revealing statistically significant syntactic patterns that distinguish higher-scoring exams according to “informational production” and lower-scoring essays according to “involved” or “interactional” production (Biber, 1988). These differences contribute to what we label emphatic generality in the lower-scoring essays, in which writers tend to foreground human actors, including themselves. In contrast, patterns in higher-scoring essays achieve what we call elaborated specificity, by focusing on and explicating specific, often abstract, concepts.These findings help uncover what is rewarded (or not) in high-stakes writing assessments and show that some students struggle with register awareness. A related implication, then, is the importance of teaching register awareness to students at the late secondary and early university level—students who are still relative novices, but are being invited to compose informationally dense prose. Such register considerations, and specific features revealed in this study, provide ways to help demystify privileged writing forms for students, particularly students for whom academic writing may seem distant from their own communicative practices and ambitions.

Despite a long debate and the accompanying call for changes in the past few decades, grammarinstruction in college English classes, according to some scholars, has remained largely “disempowering,” “decontextualized,” and “remedial” (Micciche, 2004, p. 718). To search for more effectiveand empowering grammar teaching, this study explores the use of corpora for problem-basedlearning/teaching of lexicogrammar in a college English grammar course. This pedagogy wasmotivated by research findings that (1) corpora are a very useful source and tool for languageresearch and for active discovery learning of second/foreign languages, and (2) problem-basedlearning (PBL) is an effective and motivating instructional approach. The data collected andanalyzed include students’ individual and group corpus research projects, reflection papers oncorpus use, and responses to a post-study survey consisting of both open-ended and Likert questions.The analysis of the data found the following four themes in students’ use of, and reflectionsabout, corpus study: (1) critical understanding about lexicogrammatical and broader languageuse issues, (2) awareness of the dynamic nature of language, (3) appreciation for the context/register-appropriate use of lexicogrammar, and (4) grasping of the nuances of lexicogrammaticalusages. The paper also discusses the challenges involved in incorporating corpus use into Englishclasses and offers suggestions for further research.