This qualitative, classroom-based study investigates how 24 middle school students engaged with grammar in an instructional approach that bridged “grammar in context” practices with writing transfer practices. This “transfer-minded” approach invited students to monitor and reflect on their grammar choices in writing as a way to foster metacognition; it also guided students to consider other and future writing contexts where they could use the grammar they were learning as a way to foster transfer and long-term writing development. The dissertation findings detail the role of individual factors on student engagement with grammar in this study. The findings also share how students thought about and used grammar in their writing. Finally, the dissertation reveals how students engaged with the instructional practices, helping map successful and promising areas for effective grammar instruction.

The purpose of this study was to examine the relationship between writing motivation and performance for students in grades three to five, and whether scores differed based on students’ language status, gender, or grade level. Three student language groups were included: (1) emergent bilingual students currently receiving services for English language development (ELD); (2) reclassified bilingual students who had exited ELD programs; and (3) native English-speaking students. The district administered a performance measure and a motivational measure to participating students (N = 1126). Intrinsic and self-regulatory motivation scores were significantly higher for emergent bilingual students (intrinsic M = 2.07; self-regulatory M = 2.48) and reclassified bilingual students (M = 1.99; M = 2.71) than their native English-speaking peers (M = 1.67; M = 2.30). On extrinsic motivation for writing, reclassified bilingual students (M = 2.69) scored significantly higher than both emergent bilingual students (M = 2.49) and native English speaking student (M = 2.57). Fourth and fifth graders scored significantly higher than third graders on extrinsic (M = 2.68, 2.74, 2.31, respectively) and self-regulatory motivation (M = 2.53, 2.59, 2.21 respectively), while only fourth graders scored significantly higher than third graders on intrinsic motivation (M = 1.90, M = 1.76). The only significant difference by gender was found on extrinsic motivation, where girls (M = 2.67) scored higher than boys (M = 2.53). On the district writing test, native English speaking students (M = 5.15) scored significantly higher than both reclassified (M = 4.95) and emergent bilingual students (M = 3.94). Reclassified bilingual students scored significantly higher than emergent bilingual students. In terms of grade, gender, and performance, fifth graders (M = 5.58) scored significantly higher than fourth (M = 4.56) and third graders (M = 4.31), and girls (M = 5.07) scored significantly higher than boys (M¬ = 4.64). Finally, after accounting for significant variance in district writing scores according to language status, gender, and grade (R2 = .22), the motivational incentives for writing significantly predicted an additional 1% of the variance. Findings are discussed according to the Writer(s)-Within-Community model (Graham, 2018).

Special education identification processes related to English language learners (ELs) in the United States have puzzled the field for decades. The phenomenon of referrals, the first step toward identification, is complex since it requires deciphering the root cause of students’ learning struggles—e.g., second language (L2) factors, the possibility of a learning disability (LD), or the combination of multiple other influences. To investigate the various influences contributing to learning difficulties, I centered this study on three potential sources, individual, institutional, and interpersonal. I aimed to answer, how did sociocultural influences mediate a teacher’s understanding of ELs’ competence? How did sociocultural influences mediate whether a teacher referred ELs to special education services? Using a cultural-historical theoretical approach, I sought deeper theoretical and empirical understandings into how institutional factors (e.g., tiered intervention contexts, policies), combined with other influences, mediated ELs’ referral decisions. I used a multiple parallel case study design following two fifth-grade ELs who faced the possibility of a referral. Interested in the interpersonal domain (e.g., interactions and communication among people), I zoomed in to a local process, student-teacher conferences to examine how classroom processes shaped teachers’ thoughts of students’ competence, and ultimately, referral decisions. I video-recorded teacher-student conference sessions over 14 weeks, and audio-recorded viewing sessions of the recorded conferences to understand teacher and student interpretations of learning competence. To understand how other dimensions (individual and institutional) contributed to teachers’ overall views about the student competence, I interviewed parents and school personnel, wrote observational field notes, and examined archival documents related to student learning over the entire fifth-grade year. I used inductive and iterative qualitative analytical approaches to craft the findings. My findings reaffirmed the complexity involved in finalizing ELs’ referral decisions. I found cultural factors intertwined with structural forces, driving students’ special education candidacies in divergent directions: one evaluated (LD); the other, retained. I also found the referral decisions were based on narrow understandings of learning and behaviors, lack of attention to students’ L2 needs, and faulty and overpowering structural forces which undermined teacher’s professional opinions about the referrals. These findings have implications for research, practice, and policy.

This qualitative dissertation study examines the use of shared writing about books of choice in an eighth-grade English Language Arts classroom. Drawing on data collected from 23 eighth-grade students, this study investigates how sharing writing in a classroom community impacts how students connect with a novel and how sharing writing helps to shape students’ writing practices and identity. The qualitative data collected for this study includes open-ended surveys, written reflections, interviews,teacher-researcher field notes, and examples of student work and writing. The findings of this study demonstrate the value of book choice, the benefits of peer interaction and feedback, and the usefulness of multimodal composition. These findings present ways that secondary teachers can improve both writing instruction and literature study.

An explanatory sequence mixed methods design was used to examine the effects of choice on the writing performance and motivation of college students (n = 242). The randomized control trial was followed by semi-structured interviews to determine the perceptions students (n = 20) held on the experiment as well the importance of choosing writing topics in college writing assignments. The effects of choice were tested as part of a real writing assignment that was included in nine sections of an introductory special education course. Results from hierarchical linear modeling (HLM) analyses found choice had a statistically significant negative effect on holistic writing quality, number of words written, and intrinsic writing motivation. Findings from the semi-structured interviews provided context for understanding the unexpected quantitative results.

This study examines the effects of providing persuasive writing and reading comprehension strategy training on source-based essay writing. Strategy training was administered through the use of the Writing Pal and the Interactive Strategy Trainer for Active Reading and Thinking (iSTART). The impact of both individual (writing or reading) and blended strategy training on source-based writing was investigated. A total of 261 participants completed the study; after removing incomplete and second language participants the source-based writing and system performance was assessed for 175 participants (n no instruction = 48, n iSTART =41, n Writing Pal =41, n blended =45).

Results indicated that participants who received blended strategy training produced higher quality source-based essays than participants who received only reading comprehension, writing strategy training, or no training. Furthermore, participants who received only reading comprehension or writing strategy training did not produce higher quality source-based essays than participants in the no-training control group. Time on task was investigated as a potential explanation for the results. Neither total time on task nor practice time were predictive of group differences on source-based essay scores. Analyses further suggested that the impact of strategy training does not differ as a function of prior abilities; however, training does seem to impact the relation between prior abilities and source-based essay scores. Specifically, prior writing ability was unrelated to performance for those who received writing training (i.e., Writing Pal and blended conditions), and prior reading ability was unrelated to performance for those received the full dosage of iSTART training. Overall, the findings suggest that when taught in conjunction with one another, reading comprehension and writing strategy training transfers to source-based writing, providing a positive impact on score. Potential changes to the Writing Pal and iSTART to more closely align training with source-based writing are discussed as methods of further increasing the impact of training on source-based writing.

The primary objective of this study was to develop the Perceived Control of the Attribution Process Scale (PCAPS), a measure of metacognitive beliefs of causality, or a perceived control of the attribution process. The PCAPS included two subscales: perceived control of attributions (PCA), and awareness of the motivational consequences of attributions (AMC). Study 1 (a pilot study) generated scale items, explored suitable measurement formats, and provided initial evidence for the validity of an event-specific version of the scale. Study 2 achieved several outcomes; Study 2a provided strong evidence for the validity and reliability of the PCA and AMC subscales, and showed that they represent separate constructs. Study 2b demonstrated the predictive validity of the scale and provided support for the perceived control of the attribution process model. This study revealed that those who adopt these beliefs are significantly more likely to experience autonomy and well-being. Study 2c revealed that these constructs are influenced by context, yet they lead to adaptive outcomes regardless of this contextual-specificity. These findings suggest that there are individual differences in metacognitive beliefs of causality and that these differences have measurable motivational implications.