1 of 31

Effects of a Cognitive Strategies Intervention to Improve Teacher Instruction and Students’ Argument Writing in History

Tamara Tate

University of California, Irvine

2 of 31

Meet the Team

Authors:

Steiss, Olson, Baker, Graham, Kim, Tate, Krishnan, Chung, Collins, Gilbertson, Moon, and Wang

3 of 31

The Need

Students are not writing well
Problem can’t be solved in a single subject area (ELA)
Disciplinary literacy is important (Common Core, History Social Studies Framework)
Need writing instruction that scaffolds the complexity of communicating in argumentative genres and disciplinary thinking
History teachers less prepared than ELA for this role

3

Digital Le LRA Annual Meeting | December 2024

I imagine that most of you are aware that today’s K-12 students are not writing well and that the lack of even basic writing skills is widespread. Despite the need for strong writing skills for college and career success, we are not preparing students for their futures. And even if every English teacher had the time, training, and resources to maximize students’ writing education in their English language arts classroom, we won’t prepare students for the variety of genres that they will need. We saw with the Common Core standards and the history social studies framework an articulation of the need for students to write across disciplines, in disciplinary appropriate ways. Students need to be able to use critical thinking skills, synthesize evidence across multiple sources, and marshall evidence clearly and persuasively. But history teachers have not received much training in teaching writing.

4 of 31

Study Purpose

To enhance the multiple source-based argument writing of 6th-12th grade students in history/social studies classrooms

4

Digital Le LRA Annual Meeting | December 2024

5 of 31

Background

Cognitive & sociocognitive theories of writing development
Writer[s]-within-Communities model of writing

Self-efficacy predicts writing quality

Disciplinary literacy

5

Digital Le LRA Annual Meeting | December 2024

The present study draws upon cognitive and sociocognitive theories of writing development, as well as the concept of disciplinary literacy. The cognitive process theory of writing, attributed to Flower and Hayes (1981), describes writing as a set of unique cognitive processes or strategies (e.g., idea generation, making conclusions, critiquing, or revising) that writers use during composition. Through a sociocognitive lens, effective writing instruction addresses the cognitive load writers face in complex tasks like argument writing by providing scaffolding, apprenticing novices into literacy activities, and explicitly teaching strategies with a gradual release of scaffolded support so students increasingly self-regulate and take ownership of their performance (Bruner, 1978; Englert et al., 2006; Graham et al., 2016; Vygotsky, 1986). Because writing is a cognitive and social activity, writing happens within a specific community that communicates the norms, values, tools, and practices writers take up and reproduce through their practices (Graham, 2018). Graham’s Writer[s]-within-Communities (WWC) model of writing accounts for how individuals' beliefs about their writing abilities—their self-efficacy—influence how they approach writing tasks, persist in them, and develop as writers over time (2018). Students with high self-efficacy for writing persist in difficult writing tasks, while writers with lower beliefs in their competence are more likely to disengage when challenged (Bandura, 2006; Dweck, 1999; Pajares et al., 2007). Studies have found that self-efficacy predicts writing quality, even after accounting for writing ability, grade level, and gender (Graham & Harris, 1989; Pajares et al., 1999). Because students also need knowledge of the norms and practices of distinct writing communities, like history classrooms, to understand and display the conventions for sound writing (Graham, 2018), we also use disciplinary literacy as a lens to understand student writing development. Disciplinary literacy emphasizes the relation between the norms of the writing community and individual practice by highlighting how the specific skills for communicating and norms for making knowledge claims in different subject areas vary (Moje, 2008; Shanahan & Shanahan, 2008). For example, historians use different types of data (e.g., primary sources, letters, paintings) and interpret this data using distinct cognitive strategies (e.g., sourcing, contextualization, and corroboration) compared with a scientist or literary critic. Specifically, history prioritizes making evidence-based claims about the past using various historical sources and interpretive lenses (De La Paz et al., 2017; Nokes, 2013; Nokes & De La Paz, 2023; Wineburg, 1991).

6 of 31

Background

Cognitive strategies

Explicit instruction in cognitive strategies improves student critical thinking, reading, and writing

General writing strategies
Disciplinary specific strategies

Evidence that strategy instruction is especially effective for students developing English proficiency

6

Digital Le LRA Annual Meeting | December 2024

Based on a sociocognitive view of writing development, cognitive strategies are acts of mind, or thinking tools, that readers and writers use to construct meaning (Olson, 2010). These include goal setting, tapping prior knowledge, evaluation, sourcing, and other strategies needed for meaning-making in a discipline. Explicit instruction using cognitive strategies improves student critical thinking, reading, and writing in multiple contexts and prepares students to be college and career ready (Block & Pressly, 2002; Conley, 2013; Graham et al., 2023; National Institute of Child Health and Human Development, 2000; Olson et al., 2020; Tierney et al., 1989). strategy instruction is especially effective for students developing English proficiency due to an explicit focus on language, increased exposure to academic texts, the introduction of comprehensive strategies, and multiple opportunities to affirm or correct their understandings and uses of language

7 of 31

3,400 research articles vetted

55 chosen for WWC review

15 met WWC standards

6 of the 15 were from Pathway, C3WP, & SRSD

“Teaching students to use cognitive strategies is one way to develop strategic thinking skills, ultimately helping them write more effectively” — IES Practice Guide; Recommendation # 1

8 of 31

8

Digital Le LRA Annual Meeting | December 2024

9 of 31

9

Digital Le LRA Annual Meeting | December 2024

10 of 31

10

Digital Le LRA Annual Meeting | December 2024

From September 2022 to May 2023 we met with teachers for three full-day PD workshops in the fall (18 hours), four after-school workshops in the winter/spring (8 hours), and a final full-day workshop in the spring (6 hours) for a total of 32 hours. In each of the eight PD workshops, teachers learned new instructional strategies for teaching students reading, thinking, and writing in history. Teachers were apprenticed into learning instructional moves through model, practice, and reflect cycles, reflecting a sociocognitive apprenticeship (Englert et al., 2006; Graham, 2018). Using a model inquiry unit, we engaged teachers by modeling and scaffolding instructional strategies throughout the year. After modeling specific instructional strategies (see Appendix A for details), we provided teachers time and opportunities to practice instruction and plan to implement new strategies between the end of the current workshop and before the next meeting. At the beginning of the subsequent meeting, teachers reflected on how the implementation went with their students. The structure of PD intended to gradually build teachers’ competencies in reading and writing instruction so they could effectively implement a 2-week grade-level specific reading, writing, and revision unit in the Spring term.

11 of 31

11

Digital Le LRA Annual Meeting | December 2024

The first PD workshop focused on using a cognitive strategies approach to read and analyze multiple sources, including guiding students in using a cognitive strategies “bookmark” to make meaning from historical sources. At the next workshop, teachers learned instructional strategies to help students reason with multiple sources, engage in presenting and addressing counterarguments, deconstruct a writing prompt, and argue using historical lenses. At the third workshop, teachers learned strategies for helping students compose and revise a source-based argument essay in history and engage in self-assessment and goal-setting to improve their writing over time. In the second semester, teachers attended after-school workshops on the following topics: modeling a think-aloud when revising writing, providing focused and actionable feedback to student writing, additional close reading strategies for primary sources, identifying key elements in effective source-based inquiry questions, integrating writing-to-learn activities, and analyzing text structures to aid students in reading from different genres.

12 of 31

Introducing Students to the Cognitive Strategies in Their Reader’s and Writer’s Tool Kits for Use in Your History Class

UCLAIMS, 9/20/2022

This unit is designed for teachers in the UCLAIMS project as a model of how to integrate a cognitive strategies approach into reading and writing instruction in history. You will need to apply to your own context, particularly focusing on the mid-year unit. Teachers for whom this content is relevant can teach the unit. Others can select a text set of sources for content they are currently teaching to introduce the cognitive strategies to students. This cognitive strategy practice should be reinforced throughout the school year with subsequent content.

So… The first step in the unit is to help students understand that when we read and write, we have thinking tools or cognitive strategies inside our heads that we access to construct meaning. Researchers say that when we read, we’re composing, just as when we write. What they mean is that while we read, we’re creating our own draft of the story inside our heads and as we keep reading and come across something we didn’t expect to happen or suddenly make a big discovery about what something means, we start on a second draft of our understanding.

11:30

13 of 31

Planning and Goal Setting

My purpose [for reading this source] is...
My top priority is...
I will accomplish my goal by...

14 of 31

We will need to consult additional sources to corroborate or update our initial interpretations

Corroboration

I need to check another source because…
Another source that confirms/challenges what I am reading here is...
Although this says ___, other pieces of evidence show...

15 of 31

Historical Contextualization / Perspective Taking

At the time this was written/created, people were concerned with...
From ______’s perspective, I can understand...
Given the historical context, I think ____really matters because ...

Historical Contextualization / Perspective Taking:

In both versions of the speech, Sojourner Truth takes a much more personal approach to delivering her speech than Elizabeth Cady Stanton does. Whereas Stanton says “we” and “women,” Truth says “I” and “woman.” When you engage in perspective taking, you think about the time people lived in, and the circumstances they encountered, and try to see things through their eyes. You might say, “At the time this was written/created, people were concerned with…,” “From the author’s or historical person’s perspective, I understand…,” or “Given the historical context, I think___ really matters because….”

Step into Sojourner Truth’s shoes. Given who she is, where she was, and what she was trying to accomplish, why did she tell her personal story in her speech?

1:30-1:35 partners or volunteers from whole group to respond

16 of 31

Sourcing

Because the author of this document was ____, I think their purpose was...
Because the author’s perspective is ____, we might be missing
This source is reliable/unreliable because...

Break down and model the strategy

Apply to different texts

Students apply with support

Adapt for different purposes

17 of 31

17

Digital Le LRA Annual Meeting | December 2024

Though teachers implemented disciplinary reading, thinking, and writing strategies throughout the year, during the second semester of the school year, teachers implemented an approximately 2-week inquiry unit focused on revising source-based writing through teacher instruction. This mid-year unit featured a source-based argument prompt with four sources based on grade-level content. Units were designed with pilot year teachers and the research team during the previous year to emphasize key reading, thinking, and writing skills.

Specifically, students read sources and the prompt and then wrote an initial response to the question. Students then received feedback on their writing, used a self-assessment tool, and set goals to improve their writing. Next, teachers led students through instruction-specific stages of instruction, including dissecting the prompt using a Do/What chart, comparing less effective and more effective essays written to the same prompt and identifying features of the best essays, using cognitive strategies to reread specific sources, revising key parts of the essay with lessons on academic writing or evidence use, and reflecting on their revised writing. Most teachers also had students complete an argument writing graphic organizer to help structure their writing, with some teachers having students complete this before the first draft and other teachers using it before the revised draft. There was variability in how teachers implemented the mid-year unit, which is documented in the fidelity measure discussed later. Figure X shows each grade-level inquiry unit.

18 of 31

Research Questions

What are the effects of the intervention on students’ writing achievement, and how do the effects differ for students in the classes of teachers participating in their first year of the intervention compared to those in their second year of the program?
Do the effects of the intervention differ among students of varying language proficiencies?
How does teacher fidelity to the intervention moderate the impact on student writing?
What are the effects of the intervention on students’ self-efficacy for writing and self-reported cognitive strategy use?

18

Digital Le LRA Annual Meeting | December 2024

19 of 31

Methods

Quasi-Experimental Design

A year-long professional development (PD) intervention to improve writing skills in history
Three conditions for teachers

Randomly selected teachers placed into treatment (n=12)
Randomly selected teachers placed into control (n=15)
Teachers from field trial year (volunteers) continued for a second year (n=15)

19

Digital Le LRA Annual Meeting | December 2024

The research team began a partnership with two urban, Title 1 school districts in the Southwest United States during the 2020-2021 school year. 73% of students in District A and 70% in District B received free or reduced-priced lunch (FRPL). Approximately 70% of the students in both districts were Hispanic/Latinx. The districts were linguistically diverse, with approximately 15% of the students designated as English learners (ELs), 25% designated as reclassified fluent English proficient (RFEP), and the remaining students designated as English only or initially fluent English proficient (IFEP). Language proficiency was determined by both districts using language testing and teacher referrals.

The partnership moved through three phases over time: a development phase with six teachers, a pilot phase with 22 additional teachers, and a randomized control trial (RCT) described in this article. The development phase began in the 2020-2021 school year and was significantly disrupted by the COVID-19 pandemic. The pilot phase took place the following year, and the RCT phase took place during the 2022-2023 school year with 21 teachers from these districts. District leaders reported that continued disruptions and challenges to instruction caused by the ongoing effects of the COVID-19 pandemic caused difficulties in recruiting. These challenges included student absences, teacher absences, a shortage of substitutes, and teacher stress, all of which were reported elsewhere in the U.S. (Chan et al., 2021; Dorn et al., 2022; Institute of Education Sciences, 2022; Kuhfeld et al., 2022; Siakalli et al., 2022). Further, participating districts were especially impacted given the disproportionate economic, social, and physical effect the pandemic had on communities of color (Center for Disease Control and Prevention, 2022). As a result of these issues in recruitment, the study team recruited an additional local school from a third district to participate, with seven teachers agreeing to participate, bringing the total number of teachers to 28, well under the study’s intended goal of 60 (see Table X).

20 of 31

Methods

20

Digital Le LRA Annual Meeting | December 2024

This study featured randomly assigned treatment (n=13) and control (n=15) teachers, with treatment teachers participating in a yearlong writing intervention consisting of PD and control teachers teaching writing as usual. In addition to randomly selected treatment and control groups, we also invited teachers who participated in the pilot study of the intervention to participate in a second year to evaluate the impact of an additional year of the treatment (n=15). This design allowed us to compare the effects of the intervention on students for two different groups of teachers—randomly selected teachers in their first year of PD and self-selected teachers in their second year of PD. We refer to these teachers as first-year and second-year teachers throughout this manuscript. Second-year teachers participated in a pilot phase, or test of the intervention, during the previous academic year. They attended eight PD workshops and administered pre- and posttests of student writing. The workshops these teachers attended were similar to the RCT year, but some evidence-based changes were made between the pilot and RCT years.

21 of 31

Methods

21

Digital Le LRA Annual Meeting | December 2024

22 of 31

Methods

Writing Task

Pre- and post-tests used source-based argumentative writing tasks to measure proficiency
Two 50-minute class periods to read sources and write essays

22

Digital Le LRA Annual Meeting | December 2024

All teachers administered pre- and post-writing assessments to students in their history classes.Two different prompts were used at pre- and posttest, but both had the same format, directions, and number of sources (4). The prompts required students to make an argument of causal analysis about how a major social movement succeeded by integrating evidence from primary and secondary sources to support an interpretation—key competencies required for effective argument writing in history (CCSI, 2012; Goldman et al., 2016; Monte-Sano, 2011, 2012; Wiley et al., 2020; Wiley & Voss, 1996; Wineburg, 1991). The first prompt asked: What is the most significant reason the Montgomery Bus Boycott succeeded? The second prompt asked: What is the most significant reason the Delano Grape Strike and Boycott succeeded? In order to promote elaborative argumentative reasoning and historically sound thinking, in our prompt, we asked students to consider and provide evidence for multiple reasons why the movements succeeded before arguing why their reasons were more significant. Students read the prompt and sources on Day 1. If students finished reading on Day 1, they could plan for writing by making notes or an outline for Day 2. On Day 2, students were instructed to read the writing prompts and write an essay using the sources and any notes/outlines they made the day before. All students responded to the prompt using Google docs, which they were familiar with, and typed their essays using keyboards. Students also had the option to hear the directions, prompt, and sources read aloud.

23 of 31

Methods

Writing Task

Pre- and post-tests used source-based argumentative writing tasks to measure proficiency
Two 50-minute class periods to read sources and write essays

23

Digital Le LRA Annual Meeting | December 2024

We used random sampling to select students to measure the writing outcome (holistic scoring) due to resource constraints and the cost of scoring essays. For all students who submitted pre- and post-assessments, we randomly selected 267 students from the treatment group, 267 from the returning pilot year teachers, and 266 from the control group (n=800). Then, we selected all remaining English learners (n=75) to increase the power of our analyses to answer our research question related to the effects of Intervention X on English learners. Trained evaluators scored pre- and post-writing assessments in the summer after the intervention. We trained 18 raters in a holistic rubric using anchor papers and discussion of key features of essays representing each of the six distinct categories—1 (“No Evidence of Achievement” ) through 6 (“Exceptional Achievement”). Raters were secondary literacy and history teachers or graduate students majoring in education or history. Two raters scored each essay. Essays were assigned to scorers randomly. The average agreement within one point was 94% for the evaluators (De La Paz et al., 2012; Kuhn et al., 2016; MacArthur et al., 2019; Troia et al., 2019). A third evaluator scored essays with scores that disagreed by more than one point.

24 of 31

Self-Efficacy and Cognitive Strategies

Student self report on use of cognitive strategies for writing and reading
Student self-efficacy for writing

24

Digital Le LRA Annual Meeting | December 2024

25 of 31

Methods

Analytic Method

25

Digital Le LRA Annual Meeting | December 2024

To address our first research question examining the effects of Intervention X, we used multiple regression with clustered standard errors at the teacher level because the treatment assignment was done at the teacher level (Abadie, et al., 2023). First, students’ post holistic scores were regressed on their treatment condition (Treat), controlling for their pretest holistic scores and covariates (sex, race, district, grade level special education status, and parent education) to examine the overall effect of Intervention X. Students’ EL status was added to the baseline regression model to see if the effect of the intervention differed as a function of their EL status. We then included an interaction term between students’ treatment group and EL status to examine the moderating effect of EL status. Next, regression analyses excluding control teachers were conducted to examine teacher fidelity. A latent fidelity variable for all teachers who received treatment (both year 1 and year 2 teachers) was created using available observational data, PD attendance, and fidelity in completing components of the mid-year unit.

26 of 31

Results: Writing Achievement

26

Digital Le LRA Annual Meeting | December 2024

Intervention X provides evidence that a cognitive strategies approach can improve student argument writing in history. The growth of students from second-year teachers’ classrooms resembles growth in similar quasi-experimental interventions seeking to improve students’ disciplinary writing (De La Paz et al., 2017). We see these positive effects on writing growth as attributable to a few key factors, including components of the intervention that were widely utilized across teachers: explicit cognitive strategies instruction, frequent opportunities to practice source-based inquiry, using graphic organizers and other scaffolding to facilitate the planning and organization of writing, color-coding effective and ineffective essays, self-assessment and goal setting, and using sentence frames in specific aspects of writing.

27 of 31

Results: Writing Achievement

English Learners

27

Digital Le LRA Annual Meeting | December 2024

the intervention was more successful for fluent English speakers than their English learner peers. This suggests that future iterations of curriculum and instruction should consider the need for additional scaffolding or practice opportunities for EL students who are still developing proficiency in academic language and learning the conventions of disciplinary literacy. Further research is needed to parse out these findings and suggest how to improve the intervention to reduce the gap between our EL students and their peers.

Though the intervention was less effective for ELs than EO/IFEP and RFEP students in second year teachers’ classrooms, these students still improved in their writing and would be scoring on par with their non-EL peers in the control group. Further, there was no difference between students designated as always English fluent and students reclassified as English proficient (these students were once ELs), showing the intervention worked for students equally once a baseline level of English proficiency was documented. Specific efforts were made to support English learners, including developing Spanish language versions of the cognitive strategies bookmarks, explicit instruction in the features of academic language and activities for building these skills, and scaffolding complex writing tasks through graphic organizers and a combination of talk and writing. Notably, all attempts at modifying instruction occurred at the general classroom level, so additional reading and writing courses may better serve students to focus on foundational and discipline-specific skills. A final factor to consider when examining the difference in writing growth for ELs and non-ELs is the disproportionate impact of COVID-19 on communities of color, of which virtually all the ELs in the present study were a part of (CDC, 2022).

28 of 31

Results

Controlling for student demographics, prior achievement, and English language status, we find no significant difference between student writing quality for control and treatment teachers.
However, the moderate effects of the intervention were significant (Cohen’s d = .29) for students whose teacher previously participated in a pilot year of the intervention and opted to participate for a second year.
Across treatment and returning teachers, fidelity to the intervention significantly predicted student writing improvement.

28

Digital Le LRA Annual Meeting | December 2024

In Model 1, only teachers who received two years of treatment showed a significant difference from control group teachers in student writing growth (b = .65, p < .05). There was no difference in writing growth between middle school and high school students (b = .23, p = .16). In Model 2, EL students grew .53 less than EO/IFEP students after the intervention across all conditions, controlling for pretest writing scores (p < .001). Once we introduced an interaction term, the coefficient on the EL dummy variable was no longer significant (b = -.07, p = .73), but only the interaction term between EL and Treatment Y2 was significant (b = -0.63, p < .01). The differential impact we find with respect to ELs is limited to the year 2 teachers, where the EL student growth was significantly less than the student growth by our reference group of fluent English speakers. Specifically, EO/IFEP Treatment Y2 students grew .76 points on average while ELs in the Treatment Y2 group grew .13 points (= .76 - .63) on average compared to their control group peers.

Teachers’ fidelity had a small but significant effect on the intervention effect, where 1 point increase in teacher fidelity was associated with a .063 (year 1) and .068 (year 2) point increase in student writing growth (p < . 05). To put these results into context, the difference in student writing growth between teachers with the lowest and highest fidelity score is approximately .6 to .7 points on a 6-point holistic writing scale.

29 of 31

Cognitive Strategies & Self-Efficacy

29

Digital Le LRA Annual Meeting | December 2024

Contrary to our expectations, the intervention did not improve self-efficacy for writing or knowledge of cognitive strategies for reading. Even the students from second-year teachers’ classrooms, who improved in writing overall, did not significantly improve their self-reported self-efficacy.

One potential explanation for this finding is that the students in the context of argument writing in history are still developing their beliefs about their writing capabilities. As knowledge of writing features and their progress as writers are needed for students to form efficacy beliefs, the lack of sustained writing opportunities in history for students in these districts and the relative novelty and difficulty of complex disciplinary writing may result in still evolving efficacy beliefs. In short, these students have not yet developed strong beliefs or identities as writers in history.

For the students in the classrooms of first-year teachers, it is also possible they did not improve their self-efficacy for writing or their knowledge of cognitive strategies for reading and writing due to a lack of development in writing or insufficient exposure to writing opportunities.

30 of 31

Conclusion

No significant effect

Underpowered
Covid
Second year results suggest more time might be required

History teachers need more professional development to teach writing well
Time and resources to enact intervention increase effectiveness

30

Digital Le LRA Annual Meeting | December 2024

The lack of a significant effect for the treatment teachers was contrary to our expectations. However, the underpowered nature of the RCT contributed to a difficulty in detecting a small to moderate effect, if any were present. We can not rule out an effect below .62 based on our MDES, which, given prior studies in history, seems unattainable. Such a high effect size was even more unlikely given the realities of the impact of COVID-19 on our participants—they were under-resourced, exhausted, and most students were below grade level in all subjects. Regardless of this null finding, the performance of the second-year teachers compared to the first-year randomly selected group is a notable finding that policymakers and others concerned with improving writing through improved instruction should acknowledge. We view the outsized performance of second-year teachers as being due to two factors, described further below.

History teachers need more professional development and tools to teach the cognitively demanding skills of SBAW to secondary school students. We found that student writing did improve, particularly as teachers gained additional practice in the intervention (year 2 teachers), but at least in year 2, the intervention was more successful for fluent English speakers. This suggests the need for additional scaffolding for teachers who are new to implementing writing instruction and for EL students who are still developing proficiency in academic language. Further research is needed to parse out these findings and suggest how to improve the intervention to reduce the gap between our EL students and their peers.

We also found that treatment fidelity improves student outcomes. Teachers need time and resources to enact the intervention successfully in their classes as those who do see better results. Writing, and teaching writing, are both challenging and we need to provide teachers with research-based tools, time to integrate them into their classrooms, and ways to support teachers in intervention uptake.

This study shows that research-based interventions can improve student argumentative writing, but that it takes time and fidelity to see significant results. Teachers need to incorporate the practices into their teaching, and we can not solve students’ writing challenges with brief, one-time solutions.

Second-year teachers universally mentioned in focus groups at the end of the year that they understood instructional strategies more clearly in the second year. Instructional ownership, which many second-year teachers leveraged in their modifications of curricular materials and strategies, takes time to develop (Anderson, 1997; Coburn, 2003). Underscoring this point, we noted virtually no complaints that the workshops were redundant for our second-year teachers despite many similar components in the pilot and RCT years

The intervention in this study adapted highly effective ELA strategies for the history classroom. We see that moving these practices into a different discipline greatly reduced their effectiveness. History teachers’ lack of instruction in prior writing pedagogy and practices may have reduced their effectiveness in the initial year of the intervention and required additional time to incorporate the strategies into their routine.

31 of 31

Thank you to our teachers & students!

tatet@uci.edu

31

The research reported here was supported by the Institute of Education Sciences, U.S. Department of Education, through Grant R305C190007 to University of California, Irvine. The opinions expressed are those of the authors and do not represent views of the Institute or the U.S. Department of Education.