Inclusive and Supportive Education Congress
International
Special Education Conference
Inclusion: Celebrating Diversity?
1st - 4th August 2005. Glasgow, Scotland
about the conference
programme
registration
accommodation
contact
|
Inclusive and Supportive Education Congress 1st - 4th August 2005. Glasgow, Scotland |
home about the conference programme registration accommodation contact |
Mary T. Brownell,
Anne M. Bishop,
Janette K. Klingner,
Russell Gersten
Joe Dimino,
Diane Haager,
Shailaja Menon,
Randy Penfield,
Paul T. Sindelar
For correspondence please contact: Mary Brownell, mbrownell@coe.ufl.edu
There is little dispute over the importance of teacher quality for our nation’s schools. For decades, parents have recognized its importance, competing for their child’s entrance into the “best teacher’s classroom.” Recently, policy makers have elevated its importance, mandating a highly qualified teacher in every classroom (No Child Left Behind; Public Law 107-110, 2001). Over the past decade, researchers have amassed data documenting the important role general education teachers play in fostering student achievement (Darling-Hammond, 1999; Rice, 2003; Wenglinsky, 2002). For example, value added research clearly shows a link between teachers and their students’ achievement gains. Using student gain scores over a three year period, Rivers and Sanders (1996) demonstrated that highly effective teachers could produce, on average, gains of 53 percentile points compared to the 14 percentile point gains of the least effective teachers. Although value-added research shows a link between teachers and student achievement gains, it does not provide sufficient information about those teacher characteristics, knowledge, and skills that promote good student achievement. In general education, a body of literature has emerged that sheds light on those teacher characteristics most likely to secure student achievement gains. The majority of this research involves large-scale studies of teacher characteristics that influence student achievement (Rice, 2003; Wayne & Youngs, 2003). Less research exists demonstrating connections between student achievement gains and teachers’ behaviors, knowledge, and beliefs.
The field of special education does not have the same extensive research base on teacher quality (Brownell, Ross, Colon, & McCallum, 2003). In fact, special education researchers have long overlooked the role of teachers play in students with disabilities learning, focusing instead on developing effective innovations for these students. Underlying this focus on research-based practices is the implicit assumption that quality special education teachers are those who implement effective innovations. Research on exemplary teachers, however, suggests that an exclusive focus on implementing effective innovations is not enough. Exemplary literacy teachers provide integrated instruction embedding explicit, skills-based instruction in literacy rich environments. These teachers do not appear to implement specific interventions with fidelity, but instead, draw on a variety of pedagogical practices and materials to engage students in literacy learning and improve their achievement (Haager, Gersten, Baker, & Graves, 2003; Taylor & Pearson, Clark & Walpole, 2000). For instance, Taylor and her colleagues found that the most effective teachers of students in urban settings tended to provide students with instruction in a variety of word solving strategies, rather than focusing solely on a particular explicit phonics approach. Findings from this research suggest that researchers need to better understand what defines special education teacher quality, as effective special education teachers are also likely to draw on understandings of research-based practices, but in a integrative way. Just like effective general educators, exemplary special educators are likely to integrate what they know about interventions, students, and curriculum to craft outstanding instruction for students with disabilities.
Despite the paucity of teacher quality research in special education, a small number of studies have emerged that contribute to our knowledge base in this area. Nearly two decades ago, researchers drew on process product methodology in an attempt to identify those teaching behaviors closely tied to student achievement gains. In two separate studies, researchers demonstrated that key classroom practices, such as high student engagement in academic tasks and time spent in teacher directed instruction, significantly predicted student achievement gains in reading and mathematics (Leinhardt, Zigmond, & Cooley, 1981; Sindelar, Smith, Harriman, Hale, & Wilson, 1986). More recently, researchers have attempted to understand the thinking of expert special education teachers and develop a general understanding of the characteristics that might contribute to teacher quality (Stough & Palmer, 2003). In a qualitative study of 19 teachers nominated as experts, researchers examined the online thinking of special education teachers to gain insight into the decision making process of these teachers. Specifically, the researchers interviewed, observed and videotaped teachers. Observations and videotapes were part of a stimulated recall technique that involved taping teachers during instruction and interviewing them after lessons to enable teachers to disclose their thoughts or feelings during self-selected instructional sequences. Analysis of stimulated recalls demonstrated that special education teachers’ thinking during instruction was geared to the needs of individual students. Expert teachers made decisions based on knowledge of how the student characteristics, general education curriculum and environment, and students’ behavior and state of mind might affect how students’ respond to instruction.
The only large-scale study of special education teachers involves analysis of data collected in the Special Education Elementary Longitudinal Study (SEELS) and Study of Personnel Needs in Special Education (SPeNSE). In the SPeNSE study, Carlson, Lee, & Schroll (2004) developed a teacher quality framework that they tested using factorial analyses of survey data provided by special education teachers. Through factor analyses, Carlson and her colleagues determined that certification status, self-efficacy beliefs, reported classroom practices in reading, behavior, and inclusion, and involvement in professional activities (e.g. reading professional journals) were important components of teacher quality for students with disabilities and these components could be combined to calculate a teacher quality score for individual teachers. In the SEELS study, teacher quality scores were then used to predict student achievement gains for students with disabilities (Blackorby, Lee, & Carlson, 2004). Specifically, SEELS researchers found that teacher quality scores contributed a small amount of variance to student achievement gains above and beyond that accounted for by achievement scores obtained through pretesting and socioeconomic status. Thus, the SPeNSE and SEELS analyses demonstrated that teachers with certain characteristics could make a difference in the achievement gains of students with disabilities.
Research examining the beliefs of special and general education teachers provides the most insight into one important dimension of teacher quality. More than a handful of studies have helped us to understand relationships between the beliefs of special and general education teachers and their work with students with disabilities (Jordan, Kircaaali-Iftar, & Diamond, 1993; Stanovich & Jordan, 1998; Soto & Goetz, 1998). Landrum and Kaufman (1992) found that teachers judged by their colleagues as more successful in working with students with emotional and behavioral disorders were more likely to demonstrate a high degree of self-efficacy for working with such students. Additionally, Jordan and her colleagues (Jordan, Kircaaali-Iftar, & Diamond, 1993; Stanovich & Jordan, 1998) found that general education teachers who held interventionist versus pathenogenic beliefs were more likely to engage in practices viewed as more inclusive, engaging students with disabilities in more explicit, interactive instruction.
Although the above studies provide some insights, particularly into the beliefs and practices of teachers perceived as effective, they do not tell us how dimensions of teacher quality are related to student achievement gains. The current state of teacher quality research in special education is troubling in a policy context that values student achievement gains almost to the exclusion of other outcomes of teaching and schooling. Policy makers are drawing conclusions about how to define special education teacher quality and secure it only on the basis of selected studies of general education teachers. Specifically, findings from research examining relationships between student achievement gains and teachers’ subject matter knowledge and academic ability are being used to advocate for various teacher quality initiatives and policies. For instance, in both the No Child Left Behind Act and the newly authorized Individuals with Disabilities Education Act, policymakers have used these research findings to define highly qualified general and special education teachers as those with subject matter expertise only. Additionally, the mandate to create alternative routes to the classroom for persons with bachelor’s degrees in various content areas (e.g., those with subject matter expertise) is based, in part, on this research.
We do not know if research findings used to drive policy about teacher education for general education teachers can be generalized easily to special education. Students with disabilities have unique behavioral and learning challenges that may require considerable pedagogical expertise and efficacious attitudes on the part of their special education teachers. Unfortunately, the lack of research aimed at understanding the dimensions of special education teacher quality that affect student achievement gains makes it difficult to affirm or disaffirm current assumptions about teacher quality underlying federal policy.
The special education community needs research-based information about the dimensions of teacher quality to inform policy and the conceptualization and evaluation of approaches to preservice and inservice education programs. Moreover, this information must be tied to outcomes for students with disabilities, since these outcomes are the gold standard for evaluating the effectiveness of special education. In this paper, we describe an exploratory study aimed at defining and assessing beginning teacher quality in special education. To situate this study, we review various approaches to assessing teacher quality and their potential for understanding special education teacher quality, the complexities inherent in conducting teacher quality research in special education, and how these complexities shaped our thinking about our first study aimed at tying teacher quality to students’ achievement. We then outline methodology we used to study teacher quality, what we learned from this methodology, and how we are attempting to improve future efforts to study teacher quality in special education.
Approaches to Assessing Teacher Quality: A Historical Perspective
Historically, researchers have studied teacher quality using several different approaches. Teacher characteristic research is the most prominent approach to defining teacher quality with other approaches having either fallen out of favor or being more recent additions to the research community. Research on teacher characteristics examines specific demographic characteristics or attributes of teachers and their relationship to student achievement gains. In this genre, large scale studies link years teaching experience, teacher ability, and preparation background to achievement gains on standardized group-administered tests. Specifically, teachers with more experience, higher verbal or quantitative ability, preparation in subject matter, and those fully certified in their teaching area are more likely to secure student achievement gains. Although these findings are not always consistent across studies, the preponderance of evidence suggests they are important dimensions of teacher quality (see for review, Rice, 2003; Wayne & Youngs, 2003).
Three decades ago, process product research methodology was used to better understand the relationship between certain classroom behaviors and student achievement gains. Process product research, founded in behavioral psychology, attempted to break the complex task of teaching into discrete behaviors that could be correlated with student achievement gains. Research in this tradition found correlations between specific behaviors and student achievement gains including: (a) teach classroom rules and monitor their expectations, (b) provide clear explanations and ample instructional time, (c) maximize the opportunity for students to respond during instruction and seatwork, (d) use a brisk pace to present lessons and present new material in small steps, and (e) provide regular feedback (Berliner, 1984; Christenson, Ysseldyke, & Thurlow, 1989; Englert, Tarrant, & Mariage, 1992; Good, 1979; Medley, 1978; Rosenshine, 1986; Sindelar, et al., 1986). The education community relied heavily on the findings from process product research, using findings to influence school reform agendas and evaluate teachers (Blanton et al., 2003). Unfortunately, findings from the research were oversimplified and many scholars became concerned that emerging conceptions of expert were disconnected from the subject matter taught and other dimensions of expert teaching, such as teachers’ cognition, knowledge, and beliefs. As a consequence, process product research fell out of favor and was replaced by more qualitative studies and quantitative studies of teachers’ thinking and beliefs about instruction, students, and schooling.
Over the past several decades, a considerable research base has been amassed, particularly in the area of teacher beliefs. Multiple qualitative and quantitative studies demonstrate that teachers’ beliefs about students, learning, curriculum, and teaching competence affect how they enact their practices in particular subject areas and address the needs of individual students, particularly those who struggle to learn and behave (Jordan, Kircaaali-Iftar, & Diamond, 1993; Pajares, 1992; Richardson & Placier, 2001; Stanovich & Jordan, 1998). A smaller body of literature examining teacher cognition and knowledge has helped educators understand differences in the thinking of expert versus novice teachers, and the role that pedagogical content knowledge (i.e., one’s knowledge of how to enact content learning in the classroom) plays in how teachers represent subject matter for students (Stough & Palmer, 2003; Westerman, 1991; Wilson, Shulman & Richert, 1987). More recently, researchers have focused on quantifying teacher knowledge, using two strikingly different conceptual approaches to do so. In one approach, researchers focus solely on the subject matter knowledge needed for teaching and its role in observed classroom practice and student achievement gains. In the area of reading, researchers with a pure subject matter focus have examined the knowledge of language and text need to teach children to read or decode words using multiple choice items (Moats, 1994; McCutchen, et al., 2002). Mostly, this research has established that knowledge of language and text is not powerfully related to effective reading instruction or student achievement gains (Foorman & Schatschneider, 2003; Scanlon, Vellutino, Schatschneider, Gelzheiser, & Dunsmore, 2005). The second approach to measuring subject matter knowledge has focused more on accessing the content knowledge teachers use when teaching; that is, their pedagogical content knowledge. Researchers from the University of Michigan have developed multiple-choice tests with strong psychometric properties in an attempt to quantify the pedagogical content knowledge that mathematics and reading teachers use (Phelps & Schilling, 2004; Rowan, Harrison, & Hayes, in press). Although the research on beliefs and knowledge has provided rich information about different dimensions of teacher quality, it has not shed much light on how these different dimensions of teacher quality are related to student achievement gains. Only a few studies have demonstrated relationships between teacher beliefs and knowledge and student achievement gains (Muijs, D. & Reynolds, D. 2002; Rowan, Harrison, & Hayes, in press).
The recent policy push to tie what teachers do and know to student achievement growth has resulted in a resurgence of research on classroom practice, particularly in the area of literacy. Researchers using both qualitative and quantitative observation techniques have provided considerable insight into the practices of teachers who are capable of securing better than average student achievement gains in reading (Haager et al., 2003; Pressley et al., 2001; Scanlon et al., 2005; Taylor et al., 2000). Although these studies are diverse in nature, there are surprising commonalities among their findings. Effective teachers provide explicit instruction in a variety of word identification strategies within a literacy rich environment. Students are actively engaged in ongoing, intensive instruction that is differentiated to meet their needs. Effective teachers are in touch with students’ needs, in many cases relying on student observation and assessment to inform their instruction. Additionally, effective literacy teachers create warm, supportive environments where much is accomplished.
In summary, this cumulative body of research provides substantial understanding about the dimensions of teacher quality that could serve as a framework for research in special education. Clearly, effective teachers are academically capable adults who have certain experiences and preparation backgrounds. Additionally, effective teachers practice their instruction differently than teachers who are not effective, drawing on knowledge of subject matter, effective pedagogy, curriculum, and students to craft engaging, appropriate instruction. Moreover, these teachers hold certain beliefs that not only affect the manner in which they teach, but also enable them to persist in their efforts to help the most struggling learners. Yet, how do these findings apply to those teachers working in special education? And, what are the challenges inherent in applying the methodologies of the teacher quality research base in general education to special education in ways that inform our community about the linkages between dimensions of teacher quality and student achievement?
One of the greatest challenges facing researchers interested in defining and assessing teacher quality in special education is the diverse nature of special education. Special education teachers play multiple roles, from providing direct instruction in resource rooms and self-contained classrooms to working collaboratively with general education teachers in regular classrooms. Additionally, special education teachers provide instruction across a variety of subject areas to a wide range of students, both in terms of disability type and ages of students. This complexity introduces huge sources of variance into any study of teacher quality, particularly large-scale studies attempting to include special education teachers from different disability areas, content areas, and service delivery systems. Given the diversity of the special education teaching population, it is little wonder that Blackorby and his colleagues (2004) had to use a summative teacher quality score, representing several dimensions of teacher quality, to predict even a small portion of the variance in student achievement growth. Researchers seeking to establish stronger relationships between dimensions of teacher quality and student achievement gains will have to determine ways to control some of these sources of variance while still creating studies that have some external validity.
A second challenge of comparable magnitude is the assessment of student learning. The academic progress of students with disabilities is often difficult to gauge because these students make such small gains on standardized achievement tests, particularly group-administered tests. Yet, the relationship between what teachers know and do and student achievement gains is dependent on the valid and reliable assessment of student achievement. Researchers attempting to understand teacher quality cannot use scores on group administered achievement tests often employed in large-scale research on teacher characteristics or to sample the exemplary versus nonexemplary teachers used in classroom observation research. Instead, researchers will have to employ individually-administered measures of academic growth that are not always standardized, making it difficult to compare students across different achievement levels.
A third and fourth challenge inherent in this research is the variable nature of student performance and the number of students assigned to individual teachers. Students performing in the lower tail of a normal distribution vary considerably from each other, and their performance gains may be more unpredictable because of issues with test sensitivity and motivation. After collecting data for students in this study, we found that student performance gains were quite unpredictable. This issue was exacerbated by the small numbers of students assigned to a teacher, the need to collapse students across grade levels to secure a sufficient sample size of teachers, and the need to have comparable instrumentation across students (sometimes reducing the sensitivity of the instrumentation for particular students). For instance, confidence intervals around students’ mean performance gains on oral fluency for second grade reading passages were 97 units for 12 students and 168 for 4 students. Thus, the true mean for a group of 12 students existed within a range of 97 words.
Finally, the numbers of special education teachers available to study are small in number at the individual school level. In order to have a sufficient sample of teachers, researchers must include teachers from different schools and districts. Doing so, however, introduces other sources of variability, as curriculum and service delivery models vary from district to district, and even school to school.
In response to a policy context that emphasizes highly qualified teachers and different routes for preparing these teachers, we attempted to define and assess several dimensions of beginning special education teacher quality. The research presented in this paper reflects our first attempt to understand and assess the classroom practices, knowledge, and self-efficacy beliefs of beginning special education teachers. This research effort is part of a larger study aimed at evaluating the influence of different types of teacher preparation.
To reduce some of the variability inherent in this research, we focused on beginning special education teachers providing reading instruction to students with high incidence disabilities in the 3 rd, 4 th, and 5 th grades. We also selected those teachers who provided 90 minutes or more of reading instruction weekly, as a way to provide some control over the influences of other sources of instruction. Additionally, we individually assessed the achievement of participating students using a combination of curriculum-based and standardized achievement measures, as an attempt to reliably capture the achievement gains made by this population. We borrowed or adapted teacher and student instrumentation that had been validated in previous research studies. The sections that follow provide information about the methodology we used, our results, and the conclusions we have drawn from our first attempt to understand teacher quality in special education.
Participants
Teachers . Thirty-four beginning special education from Florida (n=18), Colorado (n=8), and California n=(8) volunteered to participate. They were recruited from nineschool districts that identified beginning special education teachers who were within the first three years of teaching. Teachers were equally distributed by years of experience. Personal contacts were made to determine if the teachers fit the following profile: a) held a credential or were within a year of completing an alternative route, b) directly instructed third, fourth, and/or fifth grade students with reading disabilities, c) provided reading instruction in a resource, self-contained, or co-teaching/inclusive situation and, d) taught reading at least 90-100 minutes per week to the students in the sample. Our goal was to identify special education teachers who were designated as the primary reading teacher. This proved problematic in Colorado and California because students often received reading instruction in both the general and special education classroom.
Preparation routes varied considerably. Seven teachers received master’s degrees from a 5-year unified teacher education program, four additional teachers held master’s degrees. The majority of the beginning teachers had bachelor’s degrees either in education or were completing additional coursework for certification. Four teachers in reported participating in an alternative certification program. Twenty-six teachers were fully credentialed and eight held temporary or emergency certification. The teachers were predominately female (94%) and Caucasian (97%) and were given a monetary stipend ($200) for participation.
Students. One hundred sixty five students were recruited for participation in this study. Teachers were asked to send parent permission forms home to students identified as having a learning disability, with their major difficulty related to reading. In three classrooms in California, teachers also included students (n=12) who were not formally identified as receiving special education services but were receiving reading instruction with students with disabilities. Students with severe behavior issues or more significant cognitive disabilities were excluded from the sample. One hundred (61%) were male: 79 (49%) were Caucasian, 47 (28%) were Hispanic, 28 (17%) were African American, and 11 (6%) were other or not reported. One hundred and thirty two (80%) participated in the federal lunch program, with missing data on 23 students. Over one half of the student sample was drawn from Florida with 49% of those students in the third grade. In contrast, the remaining sample was equally dividing between California and Colorado and virtually all of the students were in the fourth and fifth grades.
Setting
Data was collected in 30 schools across the three states. Sixty three percent would be classified as high poverty urban or rural schools. Service delivery models varied between self contained (n=3), resource (n=29) and co-teaching or inclusive (n=2). Instructional time in special education for reading ranged from 120 minutes a week to 550 minutes a week. One inclusive classroom in Colorado reported 600 minutes for weekly reading instruction. In 44% of the settings, students received additional reading instruction in the general education classroom. Florida was the exception as 17 out of 18 special education teachers were the sole reading instructor for the students. The number of students observed in instructional groups within classrooms also varied dramatically. Florida teachers delivered instruction in groups that varied from 2 students to 16 students. California and Colorado teachers provided instruction smaller groups ranging from 2-7.
Curriculum varied among classrooms and states. In Florida over half of the teachers used highly structured curriculum such as Reading Mastery Plus, (Engelmann et al., 2002 ) or combined both a basal with Direct Instruction materials. Three used the district adopted basal. The majority of California teachers providing instruction from the Open Court Reading Series (Bereiter et al., 2002). Colorado teachers had a less prescribed curriculum delivery model and were observed using a variety of materials and techniques. Seven of the 8 Colorado teachers had just transitioned to the district wide Readers Workshop literacy model and special education teachers used varied materials that reflected the literacy structure seen in general education, such as multi-sensory approaches, trade-books, journal writing, and packaged programs.
Teachers . For this study, we adapted an observation instrument previously used to examine first grade instructional practices for English language learners, The English Language Learners Classroom Observation Instrument (Haager, Gersten, Baker and Graves, 2003; Gersten, Baker, Haager, & Graves, in press ). The original instrument derived content from four sources: (a) observational research on beginning reading (Anderson et al., 1979; Foorman, Francis, Shaywitz, Shaywitz, & Fletcher, 1997) , b) research on sheltered English programs (Tikunoff, Ward, van Broekhuizen, Romero, Castaneda, Lucas & Katz, 1991) , (c) research on effective teaching of at risk learners (Englert, 1984; Stanovich & Jordan, 1998), and (d) the research base on components of an effective beginning reading program (e.g., National Reading Panel, 2000). This moderate-inference instrument involves observer expertise in classroom instruction in order to make reliable judgments of instructional quality (Haager, et al., 2003; Baker, et al., in press ). The original instrument reported overall internal consistency of .89 and a range from .80 to .95 for six empirically derived subscales on the instrument. Criterion-related validity was established by correlating subscale scores with residualized growth scores in actual reading performance in 20 classrooms. Correlations were as follows (Baker, et al., in press ): Explicit Teaching/ Art of Teaching (r = .62), Instruction Geared Toward Low Performers (r = .65), Phonemic Awareness and Decoding (r = .63), Interactive Teaching (.57), Vocabulary Development (.51) and Sheltered English Techniques (.49). All correlations were statistically significant (p < .05) and there were no significant differences between the specific correlations. For a moderate-inference instrument, these correlations were satisfactory.
For the purpose of this study, the English Language Learner Classroom Observation Instrument was adapted to better align with typical special education reading instruction. We used a work-group model to adapt the instrument. This involved conducting pilot observations and then discussing our separate observations item by item. Discussion continued until we reached group consensus about modifying, deleting, or adding items. Several changes were made to the original instrument. Two changes were the most significant adaptations. Items related to English language development for English language learners were omitted because this was not the focus of our study. A section on classroom management was added due to its importance in special education. In concept, the instrument content did not change. Further adaptations included rearranging, rephrasing or merging items. Some items were rearranged in the instructional practices section to better follow the instructional cycle and incorporate adjusting for individual needs. We expanded the sections on specific reading skills of word study, comprehension and fluency and added descriptive questions on the classroom environment and structure.
The final observation protocol, Special Education Observation Instrument: Features of Effective Reading Instruction, consisted of 27 items that comprised eight subscales: Instructional Practices, General Instructional Environment, Phonological Awareness, Word Study, Fluency, Vocabulary, Comprehension, and Classroom Management. The coefficient alpha reliability for the total instrument was .92. Table 1 lists the subscales and includes a sample item from each scale. Items were rated on a four-point Likert scale, though in reality it was a seven-point scale as observers used mid-point ratings such as 1.5 or 2.5. A score of 1 represented “Low Quality” for an item and a 4 represented “High Quality.” Observers could check a box marked “Not Observed” if there was no occurrence of the item in the total observation. In addition to the items and subscales, raters were asked to provide global ratings of overall effectiveness, and effectiveness of each instructional area included in the subscales (e.g, fluency, classroom management).
Content Knowledge for Teaching Reading. All teachers completed the Content Knowledge for Teaching Reading Survey developed and validated by Phelps and Schilling (2004). This survey included a 119 items categorized into 3 subscales: (a) knowledge of content in comprehension, (b) knowledge of content in word analysis, and (c) knowledge of teaching and content in comprehension. These subscales were confirmed using confirmatory factor analysis. Individual items and their relationship to scales were also analyzed by running coefficient alphas and item response theory (IRT) reliabilities. The coefficient alphas and IRT reliabilities were above .70 in each instance except for the coefficient alpha associated with knowledge of teaching and content in comprehension.
Survey of Initial Preparation and Contextual Influences on Practice. All teachers completed a survey that examined their perceptions of their initial preparation for teaching reading, the contextual influences on their practice, their self-efficacy for teaching special education students, and their self-reported practices in teaching reading. Only the self-efficacy scale was used in this analysis.
Students . The assessment protocol for students included the Word Identification and Word Attack subtests of the Woodcock Reading Mastery Test-Revised (Woodcock, 1998). WRMT-Rnu is a revision of the 1972 test and includes updated norms with a readiness component added. Salvia and Ysseldyke (1998) report that WRMT-Rnu is one of the better instruments for diagnosing reading and evidence is favorable in terms of validity and internal-consistency reliability. Data on test-retest reliability do not exist. Measures reflecting word recognition and phonemic decoding were chosen because both constructs represent skills that promote fluent, efficient word identification, which is critical to the reading process (Adams, 1990). Further, studies have shown that children with reading disabilities who receive explicit intensive instruction in phonemic decoding skills tend to make notable progress (Torgesen, Rashotte, Alexander, Alexander, & MacPhee, 2003) and capturing student gains were paramount to this study.
Oral reading fluency was measured using curriculum-based measures (CBM), derived from research conducted by Alachua County Schools in Florida (Curriculum-based Assessment in Alachua County, Florida, 1997). This evaluation study yielded correlation coefficients ranging from .89 to .99 for reliability measures including, internal consistency, test-retest reliability and inter rater reliability. The CBM measures were derived from passages from a basal reader and Fry’s (1977) Readability Formula was used to determine the reading grade levels of passages. CBM measures of Oral Reading Fluency (ORF) were selected because a) the correlation between fluent reading of connected text and overall reading ability has consistently been supported (Fuchs, Fuchs, Hosp, & Jenkins, 2001) , b) fluency correlates with comprehension (Dowhower, 1987; Fuchs, Fuchs, & Maxwell, 1988; Jenkins, Fuchs, van den Broek, Espin, & Deno, 2003), and c) CBM is an appropriate measure to use for establishing reading growth rates for students with disabilities (Deno, Fuchs, Marston, & Shin, 2001).
The Gray Oral Reading Test (Weiderholt & Bryant, 2001)was also administered at post testing to determine its value as an outcome measure of reading comprehension for students with learning disabilities for future work.
Research Procedures
Teacher Observations . Teachers were observed on three occasions, with few exceptions. More observations were required if the instructional time was limited to 30 minutes per day. Nine trained observers entered classrooms after giving the beginning teachers a few months to adjust to new classes. Pre-interview questions were sent prior to the visit to capture the nature of the lesson to be observed. We took anecdotal field notes that were used to complete the observation instrument. A final teacher observation instrument was completed at the end of the study, establishing an overall teacher quality score and individual scores per item.
Inter rater reliability was established both between and within sites. Between site inter rater scores were established by viewing 3 videotapes and de-briefing within the group. A training tape that exemplified the behaviors identified in the observation instrument was created and a training protocol that was rubric-based described each rating level for every item was reviewed. Anchor persons were established at each site who demonstrated high inter rater scores during the group meetings. Each observer at individual sites was required to establish an inter rater score greater than 80% prior to observations. In addition, inter rater reliability was monitored throughout the year with sites sending two observers out on 16 % of the school observations. Anchors traveled to California and Colorado to monitor inter rater reliability during the school year.
Survey of Content Knowledge for Teaching Reading. (Phelps & Schilling, 2004). The survey was sent to teachers to complete toward the end of the study. Teachers were instructed to answer the questions on the survey without referring to any resources.
Influences on Practice Survey and Teacher Interviews . All teachers completed a survey that addressed the preparation, context, and self-efficacy. Only the correlations related to self-efficacy are reported in this paper. Teachers reacted to seven statements that addressed their thoughts about teaching reading. For example, teachers responded on a Likert scale to statements such, “By using my expertise, I can significantly affect my students’ achievement in reading.”
In addition, a smaller subset of teachers from Florida (n=10) and Colorado (n=11) were interviewed multiple times in order to understand the interaction between their knowledge, beliefs, preparation, and contextual support as it related to their classroom practices. The results of the survey and interviews will be reported in subsequent papers.
Students . Fourteen research assistants were trained to administer and score the student assessment. Training consisted of both group and individual sessions. Pretesting occurred 6-8 weeks from the beginning of the school year and post testing followed 21- 24 instructional weeks later. Administration of the tests required approximately 40-45 minutes.
We ran two different types of analyses to better understand the classroom practices and knowledge of beginning special education teachers. First, we ran descriptive statistics to describe how beginning special education teachers were performing overall, both in terms of their classroom practice and their knowledge for teaching reading. Second, we ran two series of correlational analyses to determine how observed classroom practice and teacher knowledge might be related to student achievement gains and each other. Such correlational analyses would help us determine how classroom practices and knowledge for teaching reading differed for more effective versus less effective teachers. Because student achievement gains are one of the most important criterion for determining a teacher’s effectiveness, they became an important criterion in this study for defining the classroom practice and knowledge of effective beginning special education teachers.
In the first set of correlational analyses, we were interested in knowing if a relationship existed between the overall quality of classroom practices and student achievement gains. Additionally, we were interested in identifying those classroom practice subscales (e.g., general instructional environment) that demonstrated the strongest relationship with overall student achievement gains. Such analyses would assist in defining the important elements of effective practice in high quality beginning special education teachers.
In the second correlational analyses, we wanted to determine if a relationship existed between knowledge for teaching reading and student achievement gains, as well as a relationship between self-efficacy for teaching students with disabilities and student achievement gains. We also wanted to determine the relationships between observed classroom practice, knowledge for teaching reading, and self-efficacy. These analyses would assist us in understanding how teacher knowledge for teaching reading and self-efficacy might influence both student achievement gains and effective classroom practice. In the narrative that follows, we describe the results of the two major groups of correlational analyses for the entire sample, as well as for each of the participating sites.
Descriptive Analysis of Beginning Special Education Teachers
We calculated means and standard deviations for all beginning special education teachers. These means and standard deviations indicated that, as a group, beginning special education teachers were competent in some areas of instruction and not others. On our rating scale, a 2.5 indicated average performance; thus, any teacher designated a 2.5 or higher on the following scales was considered to be competent. For Instructional Practices, beginning special education teachers were judged to be competent (M=2.59). For General Instructional Environment and Classroom Management, beginning special education teachers were judged to be more than competent (M=2.84, M=2.92).
In contrast, beginning special education teachers, as a group, were not as competent at teaching reading. The Mean for the Reading Subscale was below average (M=1.51), with comprehension practices approaching competence. Additionally, the standard deviations for these scores were much greater indicating considerable variability in beginning special education teachers’ ability to teach reading (See Table 2).
Correlational Analyses of Beginning Special Education Teacher Practice, Knowledge, and Student Achievement Gains
As mentioned previously, correlational analyses were run to identify those aspects of assessed teacher quality that correlated with student achievement gains and each other. The results of these analyses are described next.
Observed Classroom Practice and Its Relationship to Student Achievement Gains
Before running correlational analyses, we made several decisions about how to handle both student achievement and classroom practice data to improve the precision of the correlations. For each correlational analysis, we used median student gain scores. Median gain scores were used to establish some stability in the student data, a problem resulting from considerable variance between the performance of individual students and the small number of students taught by each teacher. Additionally, only median gain scores for the second grade CBM passages were included in the analyses, along with median gain scores for word attack and word identification. Our decision to include only second grade CBM passages was based on the numbers of students included in the data who performed at the lowest levels of achievement, particularly at the Florida site. A correlational analysis of observed classroom practice and median student gains on CBM passages revealed that 2 nd grade passages were most sensitive to the performance of students at the lower achievement levels. To make this determination, we divided students into two groups, those students who fell below the mean (M=65.93) on fall second grade CBM, and those students who fell above the mean on fall second grade CBM. For the low group, we found a significant correlation between overall scores on observed classroom practice and 2 nd grade CBM gains (r=0.304. p=0.05), and for the high group we found no significant correlation. This implies that the second grade CBM passage is most sensitive at the lower achievement levels.
Additionally, individual items on the classroom practice instrument were scored as omitted if the teacher did not provide any instruction in this area. This decision was an attempt to accurately capture the quality of special education teachers’ practices in the analyses. Special education teachers in this study often provided only a portion of the students’ reading instruction; as a consequence, assigning these teachers scores in areas for which they did not provide instruction would have been an inaccurate representation of their overall teacher quality.
Pearson r correlations were used to analyze the relationship between overall quality of classroom practices and student achievement gains on the 2 nd grade CBM passages, Word Identification subtest, and Word Attack subtest. Overall Quality of Classroom Practice was a holistic score assigned at the end of the classroom observation. Interrater reliability for this score was exceptionally high (100% inter rater agreement) and highly correlated with averaged scores across the instrument (r = .936, p = .000). The Overall Quality of Classroom Practice (OCP) was significantly correlated with gains in 2 nd grade CBM passages (r=. 461, p=. 018), but not significantly correlated with Word Identification (r=. 371, p = .062) and Word Attack (r = -.146, p = .478). Additionally, General Instructional Environment, and Classroom Management were significantly correlated with 2 nd grade CMB gains (r=. 435, p=. 026; r=. 579; p=. 002), but not with WI (r=. 252, p = .215; r=. 300, p = .135) or WA (r = -.064, p = .775; r = -.017, p = .936). Other subscales of the Classroom Practice Instrument were not significantly correlated with CMB, WA, or WI gains.
We also ran separate correlations for teachers at each of the participating sites. Because the Florida site had a much greater proportion of 3 rd graders than the other sites, we wanted to determine if trends in the relationship between overall observed classroom practice and student achievement gains would vary. To address extreme outliers due to the small numbers of teachers at each site and large variability in student gain scores per teacher, we used winsorized, median gain scores. To winsorize the distribution of teacher gain scores, we take the highest teacher gain score and set it equal to the second highest, and we take the lowest teacher gain score and set it equal to the second lowest. The result of winsorizing is to pull in the most extreme median gain scores toward the center of the distribution of median gain scores. Correlational analyses revealed some interesting differences in the trends between OCP and student gains across the different research sites. In Florida, the correlation between OCP and 2 nd grade CBM gains was r=.393; whereas the correlation between CBM and WID was r=.610. Additionally, the correlation between OCP and WA was r=.107; however, this correlation rose considerably when one teacher was dropped from the analysis, more accurately reflecting the trend in the data.
In California, the trends were quite different. Although correlations for CBM 2 nd grade gains and OCP are similar (r=.374), negative trends exist for OVP and WI (r= -.125) and WA (r= -.68). It is important to keep in mind that these trends are undoubtedly influenced by the bimodal distribution of the California data, the fact that five of these teachers provided comprehension instruction exclusively, and the minimal number of 3 rd grade students in the sample.
As a contrast to Florida and California, Colorado teachers had the highest correlations between 2 nd grade CBM gains and OCP (r=.807). The correlation between WID and OCP was fairly strong (r=.444); however, a negative correlation existed between OCP and WA (r= -.547). Unlike Florida, Colorado had few third grade students in their sample and WI may be an inappropriate measure of student growth for older students.
This series of correlational analyses demonstrates that overall quality of teachers’ classroom practice plays a role in student achievement gains, particularly gains in reading fluency. Additionally, at least for beginning teachers, classroom management practices and general instructional environment play a more influential role in reading fluency gains. Strong correlations between WID and WA (omitting one teacher) at the Florida site, where half of the sample is comprised of 3 rd grade students, suggest that classroom practice plays a stronger role in the WID and WA skills of younger and lower achieving students but not older and more competent students.
Knowledge for Teaching Reading and Self-Efficacious Beliefs: Relationship with Student Achievement Gains and Observed Classroom Practice
Knowledge for teaching reading (KTR) consisted of an overall scale score and three subscale scores (i.e., comprehension knowledge for teaching and content (CNTK), comprehension knowledge of content (CKC), word analysis knowledge of content (WAKC)). The overall scale score and the three subscale scores were not correlated with student achievement gains. However, KTR, specifically WAKC did have a relationship with OCP. The correlations between WAKC and OCP was .450 (p=.016); whereas, the correlations between CNTK and OCP (r=.247, p=.206) and CKC and OCP (r=.284, p=.143) were not significant. Thus, content knowledge for teaching reading, at least at the word analysis level plays a role in the classroom practice of special education teachers. Moreover, with larger samples, the various comprehension knowledge measures may have resulted in significant correlations with classroom practice.
Self-efficacy consisted of seven questions related to the teachers’ response to statement relating to teaching reading. The OCP and self-efficacy was not correlated (r=.-.048, p=.814). Further, self-efficacy items were not correlated with knowledge for teaching reading.
In this study we explored the relationships among the observed practices of beginning special educators who teach reading to students with LD, their content and pedagogical content knowledge, and the reading achievement of their students with LD. Our overarching purpose was to determine how best to assess teacher quality among beginning special educators. We observed all teachers at least twice (usually three times) during their language arts block and rated their performance in different domains, and asked them to complete a content knowledge survey for teaching reading (Phelps & Schilling, 2004). We operationalized “teacher quality” as teachers’ scores on our observation instrument, or rating their OCP with higher scores indicating a higher quality of classroom instruction. We hypothesized that if this were true, then the students’ in the classrooms of teachers with higher ratings would show greater achievement gains. Although numerous challenges make this type of research problematic, nevertheless our findings point to some interesting trends.
We learned that quality matters. In other words, we succeeded in capturing aspects of instruction that influenced student outcomes. The series of correlational analyses we ran showed that the students of teachers with higher scores on our observation instrument achieved at higher levels than students in classrooms where the instruction was of lesser quality, particularly in reading fluency (r=.461). This was true even though almost half of the special education teachers in our sample were not the primary provider of their students’ reading instruction and students’ initial reading levels varied substantially. Another way to think about these findings is that what special education teachers do does make a difference, even given the limited amount of time some of them actually spend instructing their students.
Interestingly, it was in Colorado where seven of the eight teachers provided supplemental instruction in a resource program that we had the highest correlations between 2 nd grade CBM gains and scores on the observation instrument (r=.807). We can only speculate why this was the case. Was the total number of minutes of reading instruction higher than at other sites? That was true with California, but not Florida. Was the supplemental instruction more intensive, and more focused on children’s needs? This instruction was provided in smaller groups (2-5) than in Florida; perhaps, this was a contributing factor. Also, unlike in Florida, where half of the student sample was comprised of third-graders, in Colorado there were only a few. Because our sample only included eight teachers in Colorado, we cannot have confidence in these findings. However, they do point to an interesting trend worthy of future research.
Additionally, we found that classroom management and the general instructional environment impacted reading fluency gains more than specific components of reading instruction. General Instructional Environment and Classroom Management were significantly correlated with 2 nd grade CBM gains (r=.435, p=.026; r=.579, p=.002, respectively). Other subscales of the Classroom Practice Instrument were not significantly correlated with CBM gains.
We also noted that, in general, our beginning teachers were more sophisticated in their general instructional practices and classroom management than they were in specific reading practices. That is, their mean ratings were higher in these areas (about average or slightly higher) than the mean rating for reading-specific pedagogical practices, which was substantially below average. Thus, it appears that it is their relatively strong general instruction that is contributing to student achievement gains. It would be interesting to see if expert reading teachers (i.e., those with high scores on the reading items), would achieve even greater student achievement gains.
In some ways our work, and these findings, are similar to the process product literature. Like our predecessors 20 years ago (Brophy & Good, 1986; Leinhardt, Zigmond, & Cooley, 1981; Sindelar et al., 1986), we also established correlations between what classroom teachers do (processes) and student achievement gains (products). We too found that classroom management and general instruction features such as quick pacing and high levels of student engagement influenced outcomes. Yet our research differs from the process product research in that we attempted to more adequately capture the complexity of teaching and the roles of content knowledge and pedagogical content knowledge (Blanton et. al, 2003).
Although we saw "glimpses of expertise" in teachers’ instructional practice, especially in the general pedagogical domain, most teachers still struggled with aspects of reading instruction. For example, although most teachers regularly taught word-study using isolated or decontextualized words, we rarely saw teachers provide students with the opportunity to practice reading these (or similar) words in connected text. Also, teachers were generally weak in building background knowledge, making connections across and within lesson or with students’ experiential knowledge, and incorporating students’ responses in discussions. Although some teachers taught comprehension strategies, we saw less prompting and cueing to apply strategies. Teachers’ understanding of how to facilitate students’ comprehension seemed incomplete. As noted by the National Reading Panel (2000), teaching comprehension strategies is notoriously challenging and may require extensive formal instruction.
Accurately assessing teachers’ content knowledge and pedagogical content knowledge in reading is challenging. Although neither the overall scale score nor the three subscale scores on our knowledge survey correlated with student achievement gains, we do not believe this means that knowledge does not matter, but rather that we need to continue to develop more sensitive measures (and increase our sample size). Yet teachers’ knowledge, specifically their knowledge of word analysis, did correlate with their practice (as assessed with our observation instrument; .450, p=.016). Thus, content knowledge for teaching reading, at least at the word analysis level, influences the classroom practice of beginning special education teachers. This finding is consistent with research by Moats (1994, 1999), who emphasizes the importance of teachers having sufficient knowledge of language structure to provide effective instruction.
Challenges
The challenges are numerous in conducting this type of research. We noted tremendous across site and within site variation, in program type, in the number of minutes of literacy instruction students received, in how many different teachers were responsible for a given student’s literacy instruction, in class size, in the focus of instruction (e.g., word study or comprehension), in curricular materials, and in students’ initial reading levels.
Because there was so much variability within and across sites, and small numbers of students for some teachers, we made decisions about how to handle both student achievement and classroom practice data. To improve the precision of the correlations we used winsorized median scores. We only used 2 nd grade CBM scores as these were the most sensitive to growth among our lowest readers.
Originally we had hoped to only include special education teachers who were considered the primary reading teacher for their students with LD in this study. In Florida, where all but one special education teacher was the principle literacy instruction provider, this was feasible. However, as it turned out, this was not possible at our other sites. In Colorado and California most students received reading instruction in both their general and special education classrooms, with the special education teacher providing supplemental instruction only (with the one notable exception in Colorado mentioned above). That many special education teachers were not their students’ primary reading teachers introduced a great deal of unexplained variance because we did not collect information about the quality of instruction in general education classrooms (we only accounted for the total number of minutes of instruction they received). Nor did we find out exactly what students were taught during their time in general education (though we did ask what curricula were used). Although the special education teacher might have focused exclusively on word-study, the student most likely received comprehension instruction in the regular classroom. For this reason when a special education teacher did not provide instruction in a given area, we considered this “missing” data rather than a “zero” on our observation instrument. To truly account for all of the instruction received by students with LD, we would have needed to observe in every setting in which they were taught. Yet there was no feasible way for us to do this because students came from many different classrooms (e.g., the 4 or 5 students a teacher worked with in a given hour could be from 3 or 4 different homerooms), and we did not have the personnel to conduct this many observations.
Curricula varied across classrooms and states. It was more prescripted in Florida, where over half of the teachers used Reading Mastery Plus (Englemann et al., 2002), and in California, were they relied on the Open Court Reading Series (Bereiter et al., 1995), than in Colorado. Colorado teachers were observed using a variety of materials and techniques.
Also, service delivery models varied. The majority of our teachers taught in resource rooms (84%), while only a few taught in self-contained or inclusive classrooms. Instructional time in special education for reading ranged from 120 minutes a week to 550 minutes a week. Students in the inclusive classroom in Colorado purportedly received the most instruction, with 900 minutes a week in language arts (including reading, writing, and “skills”).
The variability in students’ entry skills was dramatic. Students’ learning needs also varied. To some extent what special education teachers focused on with their students with LD was dictated by the goals on their IEPs, and these varied. For most students the primary goal was improved word identification, though others’ goals related more to comprehension.
Another challenge was finding appropriate and sensitive measures of student growth in reading, particularly given the low reading levels of so many of the students, especially in Florida. In the fall we had assessed some students’ reading comprehension with the WJ-Comprehension but found it to be problematic with so many low readers. Thus we opted to use a different comprehension measure in the spring, the Gray Oral Reading Test (Weiderholt & Bryant, 2001), but without pretest data the utility of this data was limited.
Limitations
This study has its limitations in addition to the challenges already described. First, our sample size was too small for us to have confidence in our findings, or to allow us to run all of the analyses we would have liked. With a larger dataset we would run a multiple regression or use hierarchal linear modeling to determine the amount of variance attributable to different factors and to examine across site and within site variance.
Because almost all of our special education teachers taught in resource rooms, we suspect that our results may not generalize to teachers in other types of settings (e.g., inclusive general education classrooms). In fact, the one teacher who taught in an inclusive program in Colorado was quite different from the other Colorado teachers, emphasizing comprehension more and word-study less. She was one of our highest-rated teachers, and had the highest correlations with students’ CBM and Word ID scores. Yet she differed from other teachers in the Colorado sample not only in how and what she taught, but also in that she was her students’ only reading teacher, her students had 180 minutes of language arts instruction a day, and her students with LD started out with higher scores than many of the students in resource rooms. Thus it is difficult to know just how to interpret her students’ superior gains.
Implications for Future Research and Practice
Our research was conducted with teachers in their first three years of teaching; thus our results cannot be generalized to all special education teachers. Future research is warranted that includes veteran teachers in the sample so that the beliefs, knowledge, and practices of beginning teachers can be compared with those of more experienced teachers. We are conducting such a study this year.
We need more refined knowledge measures for special education teachers. The knowledge survey we used was developed to capture the knowledge of elementary teachers. Yet the nature of special education teachers’ knowledge may be somewhat different, perhaps more diagnostic in nature, and we need measures that can determine the ways in which special educators and general educators’ knowledge might be distinct. Some of the items seemed to capture aspects of literacy instruction that represented a higher level than most special education teachers would ever teach. Also, there seemed to be an insufficient number of questions that tapped into pedagogical content knowledge, particularly related to word analysis. A knowledge measure developed specifically for special education would also be important for large-scale studies of preparation and professional development. Time intensive classroom observations that require sophisticated observers are unrealistic for comparative studies that involve hundreds and even thousands of teachers.
Overall, beginning special education teachers have a fair amount of knowledge about teaching reading; however, they do not necessarily know how to operationalize it, suggesting a need for further research in this area. Why do some teachers apply their knowledge better than others? Are there differences in their preparation or induction into the field? Perhaps those teachers who were better able to translate their knowledge into practice were those who had been provided with more opportunities to do this in their teacher education programs. We believe this finding points to a need for continued support in this area at pre-service and in-service levels (Brady & Moats, 1997; Hoffman & Pearson, 2000; McCutchen & Berninger, 1999; Moats, 1999).
Conclusion
We have made some progress towards determining how to assess “quality” among beginning special education teachers. Our observation instrument was sensitive enough to pick up aspects of instruction that correlated with student achievement in reading even though there was a great deal of variability across and within our sites at many levels. As we continue with this line of research we will continue to refine our instruments and procedures.
The beginning special education teachers in our study provided an interesting portrayal of emerging expertise. Teachers were more effective in general aspects of pedagogy than they were in teaching reading. We suspect that reading may not have been sufficiently emphasized in many of their special education teacher education programs. We also believe that to become a skillful reading teacher takes time, experience, and on-going professional development. Although we considered a few of our teachers to be above average or even excellent in teaching specific components of reading, we rated most as below average. These findings have important implications for teacher education as well as for school districts in term of the types of support and professional development provided to beginning teachers.
Adams, M. J. (1994). Beginning to read: Thinking and learning about print (1st MIT Press paperback ed.). Cambridge, Mass.: MIT Press.
Bereiter, C., Adams, M., Adcock, I., Brown, A., Campione, J., Carruthers, I., et al. (2002) SRA Open Court Reading. DeSoto, TX : SRA/McGraw-Hill.
Berliner, D. C. (1984). The half-full glass: A review of research on teaching. In P. L. Hasford (Ed.), Using what we know about teaching (pp. 51-77). Alexandria, VA: Association for Supervision and Curriculum Development.
Blackorby, J. Lee, H., & Carlson, E. (2004). Teacher quality as a predictor of achievement for students with disabilities. Manuscript in preparation.
Blanton, L., Sindelar, P., Correa, V., Hardman, M., McDonnell, J., & Kuhel, K. (2003). Conceptions of beginning teacher quality: Models of conducting research. (COPSSE Document Number RS-6). Gainesville, FL: University of Florida, Center on Personnel Studies in Special Education.
Brady, S., & Moats, L. (1997). Informed instruction for reading success: Foundations for teacher preparation . Position paper. Baltimore, MD: International Dyslexia Association.
Brophy, J.E. and T.L. Good. 1986. Teacher behavior and student achievement. In Handbook of Research on Teaching, Third Edition, ed. M.C. Wittrock, 328-357. New York: MacMillan.
Brownell, M., Ross, D., Colon, E., & McCallum, C. (2003). Critical features of special education teacher preparation: A comparison with exemplary practices in general teacher education. (COPSSE Document Number RS-4). Gainesville, FL: University of Florida, Center on Personnel Studies in Special Education.
Carlson, E., Lee, H., & Schroll, K. (2004). Identifying attributes of high quality special ß education teachers. Teacher Education and Special Education, 27, 350-359.
Christenson, S. L., Ysseldyke, J.E., & Thurlow, M.L. (1989). Critical instructional factors for students with mild handicaps: An integrative review. Remedial and Special Education, 10, 21-31.
Curriculum-based Assessment in Alachua County, Florida: Vital Signs for Student Progress. (1997). Gainesville, FL: Exceptional Student Education, School Board of Alachua County.
Darling-Hammond, L. (1999). Teacher quality and student achievement: A review of state policy evidence. University of Washington Center for the Study of Teaching and Policy. Retrieved December 18, 2003 from http://depts.washington.edu/ctpmail/PDFs/LDH_1999.pdf
Deno, S.L., Fuchs, L.S., Marston, D., & Shin, J. (2001). Using curriculum-based measurement to establish growth standards for students with learning disabilities. School Psychology Review , 30, 507-524.
Dowhower, S. (1987). Aspects of repeated reading on second-grade transitional readers in fluency and comprehension. Reading Research Quarterly, 22, 389-406.
Engelmann, S., Arbogast, A., Bruner, E., Lou Davis, K., Engelmann, O., Hanner, S., et al. (2002) SRA Reading Mastery Plus. DeSoto, TX: SRA/McGraw-Hill.
Englert, C. S. (1984). Effective direct instruction practices in special education settings. Remedial and Special Education, 5, 38–47.
Englert, C. S., Tarrant, K. L., & Mariage, T. V. (1992). Defining and redefining instructional practice in special education: Perspectives on good teaching. Teacher Education and Special Education, 15, 62-86.
Foorman, B. R., Francis, D. J., Shawitz, S. E., Shaywitz, B., & Fletcher, J. M. (1997). The case for early reading intervention. In Blachman (ed.), Foundations of reading acquisition: Implications for intervention and dyslexia (pp. 103-115). Mahwah, NJ: Lawrence Earlbaum Associates.
Foorman, B. R., & Schatschneider, C. (2003). Measurement of teaching practices during reading/language arts instruction and its relationship to student achievement. In S. Vaughn and K. L. Briggs (Eds.), Reading in the classroom: Systems for the observation of teaching and learning (pp. 1-30). Baltimore: Paul H Brooks Publishing.
Fry, E. (1997). Fry’s readability graph: Clarification, validity, and extension to level 17. Journal of Reading, 21 , 242-252.
Fuchs, L.S., Fuchs, D., Hosp, M.K., & Jenkins, J.R. (2001). Oral reading fluency as an indicator of reading competence: A theoretical, empirical, and historical analysis. Scientific Studies of Reading, 5, 239-256.
Fuchs, L. S., Fuchs, D., & Maxwell, L. (1988). The validity of informal measures of reading comprehension. Remedial and Special Education, 9, 20-28.
Good, T. L. (1979). Teacher effectiveness in the elementary school. Journal of Teacher Education, 30, 52-64.
Haager, D., Gersten, R., Baker, S., & Graves, A. (2003). The English-language learner classroom observation instrument: Observations of beginning reading instruction in urban schools. In S. Vaughn & K. Briggs (Eds.), Reading instruction observation tools. Baltimore: Paul H. Brookes.
Hoffman, J., & Pearson, P.D. (2000). Reading teacher education in the next millennium: What your grandmother’s teacher didn’t know that your granddaughter’s teacher should. Reading Research Quarterly, 35, 28-44.
Jenkins, J.R., Fuchs, L.S, van den Broek, P., Espin, C., & Deno, S.L. (2003). Sources of individual differences in reading comprehension and reading fluency. Journal of Educational Psychology, 95, 719-729.
Jordan, A., Kircaali-Iftar, G., & Diamond, C. T. P. (1993). Who has a problem, the student or the teacher? Differences in teachers' beliefs about their work with at-risk and integrated exceptional students. International Journal of Disability, Development and Education, 40, 45-62.
Jordan, A., Lindsay, L., & Stanovich, P. (1997). Classroom teachers' interactions with students who are normally achieving, at-risk and exceptional. Remedial and Special Education, 18, 82-93.
Landrum, T. J., & Kaufman, J. M. (1992). Characteristics of general education teachers perceived as effective by their peers: Implications for inclusion of children with learning and behavioral disorders. Exceptionality, 3, 147-163.
Leinhardt, G., Zigmond, N., & Cooley, W. (1981). Reading instruction and its effects. American Educational Research Journal, 18, 343-361
McCutchen, D., & Berninger, V. W. (1999). Those who know, teach well: Helping teachers master literacy-related subject-matter knowledge. Learning Disabilities Research & Practice, 14 , 215-226.
McCutchen, D., Abbott, R.D., Green, L., Beretvas, S., Cox, S. Quiroga, T., Potter, N., & Gray, A. (2002). Beginning literacy: Links among teacher knowledge, teacher practice, and student learning. Journal of Learning Disabilities, 35, 69-87.
Medley, D. M. (1978). Research in teacher effectiveness: What it is and how it got here. Journal of Classroom Interaction, 13, 16-21.
Moats, L.C. (1994). The missing foundation in teacher education: Knowledge of the structure of spoken and written language. Annals of Dyslexia, 44, 81-104.
Moats, L. (1999). Teaching reading IS rocket science: What expert teachers of reading should know and be able to do, Item No. 372. Washington, DC: American Federation of Teachers.
Muijs, D., & Reynolds, D. (2002). Teachers’ beliefs and behaviors: What really matters? Journal of Classroom Interaction, 37, 2, 3-15.
National Institute of Child Health and Human Development. (2000). Report of the National Reading Panel. Teaching children to read: an evidence-based assessment of the scientific research literature on reading and its implications for reading instruction: Reports of the subgroups (NIH Publication No. 00-4754). Washington, DC: U.S. Government Printing. No Child Left Behind; Public Law 107-110, 201.
Pajares, M. F. (1992). Teachers' beliefs and educational research: Cleaning up a messy construct. Review of Educational Research, 62, 307-332.
Phelps, G. & Schilling, S. (2004). Developing measures of content knowledge for teaching reading. Elementary School Journal, 105, 31-48.
Pressley, M., Wharton-McDonald, R., Allington, R., Block, C.C., Morrow, L., Tracey, D., Baker, K., Brooks, G., Cronin, J., Nelson, E., Woo, D. (2001). A study of effective first-grade literacy instruction. Scientific Studies of Reading, 5(1), 35-58.
Rice, J.K. (2003). Teacher quality: Understanding the effectiveness of teacher attributes. Washington, DC: The Economic Policy Institute.
Richardson, V., & Placier, P. (2001). Teacher change. In AERA's Handbook of Research on Teaching (4th ed., pp. 905-947). Washington: American Educational Research Association.
Rivers, J.C. & Sanders, W.L. (1996). Research project report: Cumulative and residual effects of teachers on future student academic achievement. University of Tennessee Value-Added Research and Assessment Center.
Rosenshine, B. (1986). Synthesis of research on explicit teaching. Educational Leadership, 43, 60-69.
Rowan, B., Harrison, D., & Hayes, A. (in press) Using instructional logs to study mathematics curriculum and teaching in the early grades. Elementary School Journal.
Salvia, J., & Ysseldyke, J. E. (1998). Assessment. Boston, MA: Houghton, Mifflin.
Scanlon, D. M., Vellutino, F., R., Schatschneider, C. Gelzheiser, L., & Dunsmore, K. (2005, February). Instructional and teaching characteristics of kindergarten teachers whose at-risk students make the largest and smallest gains. Presentation at the Pacific Coast Researchers Conference, San Diego, CA.
Sindelar, P. T., Smith, M. A., Harriman, N. E., Hale, R. L., & Wilson, R. J. (1986). Teacher effectiveness in special education programs. Journal of Special Education, 20, 195-207.
Soto, G., & Goetz, L. (1998). Self-efficacy beliefs and the education of students with severe disabilities. The Journal for the Association of Persons with Severe Handicaps, 23, 134-143.
Stanovich, P. J., & Jordan, A. (1998). Canadian teachers' and principals' beliefs about inclusive education as predictors of effective teaching in heterogeneous classrooms. The Elementary School Journal, 98, 221-238.
Stough, L. & Palmer, D. (2003). Special thinking in special settings: A qualitative study of expert special educators. The Journal of Special Education, 36, 206-222.
Taylor, B., Pearson, D., Clark, K., & Walpole, S. (2000). Effective schools and accomplished teachers: Lessons about primary-grade reading instruction in low-income schools. The Elementary School Journal, 101, 121-165.
Tikunoff, W.J., Ward, B.,A., van Broekhuizen, L.D., Romero, M., Castaneda, L.V., Lucas, T., & Katz, A. (1991). Final report: A descriptive study of significant features of exemplary special alternative instructional programs. Los Alamitos, CA: Southwest Regional Educational Laboratory.
Torgesen, J., Rashotte, C., Alexander, A., Alexander, J., & McPhee, K. (2003). Progress toward understanding the instructional conditions necessary for remediating reading difficulties in older children. In B. Foorman (Ed.), Preventing and Remediating Reading (pp. 275-297). Timonium, MD: York Press, Inc.
Wayne, A.J. & Youngs, P. (2003). Teacher characteristics and student achievement gains: A review. Review of Educational Research, 73, 89-122.
Weiderholt, J. L. & Bryant, B. R. (2001). Gray Oral Reading Test 4. Austin, TX: Pro-Ed.
Wilson, S., Shulman, L., & Richert, A. (1987). "150 different ways" of knowing: Representations of knowledge in teaching. In J. Calderhead (Ed.), Exploring teachers’ thinking (pp. 104-124). London: Cassell.
Woodcock, R. W. (1998). Woodcock Reading Mastery Tests Revised- Nu. Circle Pines, MN: American Guidance Service.
home . about the conference . programme . registration . accommodation . contact
 |
 |
 |
 |
 |
 |