Voices from Student Teachers in New York : The Persistence of a Subtractive Experience of the edTPA as a Licensure Exam for Initial Certification

This article explores how candidates’ experiences in multiple initial certification programs within a single School of Education evolved over the first three semesters of New York’s implementation of edTPA as a requirement for initial licensure. The data reviewed included primarily surveys and interviews of teaching candidates, framed by critical perspectives on accountability, teacher performativity, and constructivist theories on learning to teach. Results suggest candidates’ perceptions of program alignment and edTPA benefits improved while several challenges persisted. These included the lack of mentor teacher knowledge about the edTPA as well as disconnects between candidates’ edTPA scores and local program evaluations. Additionally, student teachers’ perceptions of a subtractive experience of the Voices from student teachers in New York 2 edTPA continued in spite of improved perceptions about the benefits of the exam and across all scoring levels, programs, and semesters. A discussion of these results considers the implication of policies that reify quality, compel performance management, and contribute to values conflict for candidates in the particularly unique developmental moment of learning to teach. These voices of student teachers, often underrepresented in the research on edTPA, urge a reconsideration of the policy on teacher performance assessments in terms of how such policies impact the experiences of those learning to teach.

edTPA continued in spite of improved perceptions about the benefits of the exam and across all scoring levels, programs, and semesters.A discussion of these results considers the implication of policies that reify quality, compel performance management, and contribute to values conflict for candidates in the particularly unique developmental moment of learning to teach.These voices of student teachers, often underrepresented in the research on edTPA, urge a reconsideration of the policy on teacher performance assessments in terms of how such policies impact the experiences of those learning to teach.Keywords: Teacher education; teacher preparation; student teaching; teacher performance assessment Voces de estudiantes docentes en Nueva York: La perseverancia de una experiencia sustractiva del edTPA como un examen de licenciatura para la certificación inicial Resumen: Este artículo explora cómo las experiencias de los candidatos en varios programas de certificación en de una Escuela de Educación se evolucionaron durante los primeros tres semestres de la implementación de edTPA en Nueva York como requisito para la licencia inicial.Los datos incluyeron encuestas y entrevistas a candidatos a la enseñanza, enmarcados por perspectivas críticas sobre la responsabilidad, la ejecución docente y las teorías constructivistas sobre el aprendizaje de la enseñanza.Los resultados sugieren que las percepciones de los candidatos sobre la alineación del programa y los beneficios de edTPA mejoraron mientras persistían varios desafíos, incluyendo la falta de conocimiento de los maestros mentores sobre el edTPA, así como las desconexiones entre los puntajes edTPA de los candidatos y las evaluaciones de los programas locales.Además, las percepciones de los maestros estudiantes sobre la experiencia sustractiva del edTPA continuaron a pesar de las percepciones favorables sobre los beneficios del examen y en todos los niveles de calificación, programas y semestres.Una discusión de estos resultados considera la implicación de las políticas que representan calidad, motivan la administración de rendimiento y contribuyen a los conflictos de valores para los candidatos en el momento de aprendizaje particularmente singular de aprender a enseñar.Estas voces de los profesores estudiantes urgen una reconsideración de política sobre las evaluaciones del rendimiento docente en términos de cómo estas polizas impactan las experiencias de aquellos que aprenden a enseñar.Palabras-clave: Educación docente; preparación del maestro; enseñanza de estudiantes; evaluación Vozes do ensino de estudantes em Nova York: a perseverança de uma experiência subtrativa do edTPA como exame de bacharel para a certificação inicial Resumo: Este artigo explora como as experiências de candidatos em vários programas de certificação em uma Escola de Educação evoluíram durante os primeiros três semestres da implementação do edTPA em Nova York como requisito para o licenciamento inicial.Os dados incluíram pesquisas e entrevistas com candidatos para o ensino, enquadradas por perspectivas críticas sobre responsabilidade, desempenho docente e teorias construtivistas sobre o aprendizado do ensino.Os resultados sugerem que as percepções dos candidatos sobre o alinhamento do programa e os benefícios do edTPA melhoraram, enquanto vários desafios persistiram, incluindo a falta de conhecimento dos professores mentores no edTPA, bem como as desconexões entre os escores edTPA dos candidatos e as avaliações dos programas locais.Além disso, as percepções dos alunos sobre a experiência subtrativa da edTPA continuaram apesar das percepções favoráveis sobre os benefícios do exame e em todos os níveis, programas e semestres.Uma discussão sobre esses resultados considera o envolvimento de políticas que representam qualidade, motivam o gerenciamento de desempenho e contribuem para o valor dos conflitos para os candidatos

Introduction
For the last 40 years, the landscape of K-12 education has experienced dramatic shifts towards an emphasis on technical and managerial solutions to persistent problems plaguing public education (Apple, 2001;Ravitch, 2010).Many of these dilemmas have to do with student underperformance and poverty across communities increasingly segregated by race and income.To address these concerns, a group of reformers has advocated implementation of increasingly restrictive and high-stakes standardized exams alongside a call to raise standards to enable students to be college and career ready upon high school graduation.
This call to raise K-12 standards is linked to an agenda to professionalize teaching and teacher education, but is driven by ideology as much as evidence (Cochran-Smith & Fries, 2001).Such professionalization, often promoted by teacher educators and education scholars, is intended to raise the bar for recruitment, retention, and tenure of teachers.Critics, however, contend that such policies may intensify traditional and narrowed conceptualizations of curriculum and pedagogy rather than increase professional autonomy (Apple, 2001).Nevertheless, a profound shift from focusing on the inputs to the outcomes of teacher education has marked this era (Cochran-Smith, 2001) and paved the way for the nationalization of teacher performance assessment in the form of the edTPA.
Touted by proponents as an assessment "for the profession, by the profession," the aims of the edTPA are both educative and evaluative in identifying candidates' readiness to teach (Stanford Center for Assessment, Learning, and Equity [SCALE], 2015b, p. 4).The assessment involves three core tasks of planning, instruction, and assessment.The submission involves student work analysis with extensive written commentary and a video clip of instruction.With 749 institutions from at least 40 states now participating, the edTPA is the pre-eminent assessment of pre-service teacher quality in the nation (see Participation Map at http://edtpa.aacte.org/state-policy).
Based on similar assessments for novice teachers implemented in California and Connecticut and modelled on the National Board process for expert teachers, the edTPA was initially conceptualized by teacher educators who desired to create a more authentic assessment of teaching (Darling-Hammond & Hyler, 2013).Those prominent academics, who later became associated with the development of the edTPA as a national exam, presented research on early teacher performance assessments to suggest the benefits of teacher performance assessment in terms of impacts on preservice teacher learning and program change (Darling-Hammond, Newton, & Wei, 2010;Pecheone & Chung, 2006).
After a pilot, the development of the edTPA -and subsequent ownership of its contentwas moved to the Stanford Center for Assessment, Learning, and Equity (SCALE).As edTPA went national, SCALE contracted with Pearson Education, Inc., to administer the exam, collect and store related data, and coordinate scoring.According to Pearson, the intention was that the edTPA would provide evidence for teacher licensure decisions to determine whether "new teachers are ready for the job" (see edTPA at http://www.pearsonassessments.com/teacherlicensure/edtpa.html).Since its introduction as a national exam, developers and advocates have sought to establish that the edTPA design is based on research around such key scholarly constructs as constructivism (Sato, 2014), teacher effectiveness (SCALE, 2015a), and readiness to teach (SCALE 2015b).SCALE bibliographies cite limited pre-service research studies signaling a research base more grounded in the context of experienced teachers.Indeed, those few studies involving pre-service teachers typically involved teacher performance assessment as a program requirement, not as an individual exam for licensure as would become the case in New York (Clayton, 2014).
Additional controversies have emerged as the edTPA gained national prominence.Some involve the commercialization of the exam, its associated costs for candidates, the standardization of teacher quality, and a loss of autonomy for teacher educators and programs (Au, 2013;Dover & Schultz, 2016).In spite of these concerns, the codification of state policies continues apace.Sixteen states have instituted or are in the process of enacting policies that require or offer the option of participating in the edTPA (see Participation Map at http://edtpa.aacte.org/state-policy).Currently, five of these states require the edTPA as a part of the decision to award teacher licensure for initial certification with several more having plans to adopt similar policies in the near future.In analyzing policy, Cochran-Smith, Piazza, and Power (2012) employed a multi-dimensional framework, part of which focuses on "policy in practice" and "impact and implementation" (p.9).To get a sense of these levels, this article examines the dynamics of a policy in practice, through the lived experiences of teaching candidates, during the early implementation of edTPA state policy in New York.

EdTPA Implementation in New York
Among a small group of five states that have adopted edTPA as a requirement for teacher licensure, New York stands out in its rapid implementation of a high-stakes requirement for licensing individual teaching.There is little ambiguity in the adopted policy: edTPA national cut scores were adopted with no phase-in period, mastery levels were identified to be reported publicly by institution, and, after just one year of field testing, the policy was implemented for all initial certification candidates.In contrast to other states with consequential policies but four years of field testing (Reagan, Schram, McCurdy, Chang, & Evans, 2016), New York's rapid implementation led to a series of subsequent regulations to establish a "safety net" for those who failed the edTPA, allowing candidates to take the previous standardized certification exam of pedagogical knowledge for up to four years after initial implementation.Although national edTPA results in 2015 show a pass rate of 71%, the national reporting is unreliable given that states set different cut scores (SCALE, 2016).In New York, where the passing score is considerably higher, the state's pass rate was 80% in 2014, which, although it exceeds national averages, is lower than Washington's 98% pass rate, which had set a much lower cut score (Meuwissen & Choppin, 2017).
While New York's implementation has been regularly critiqued (Cochran-Smith et al., 2016;Reagan et al., 2016), it remains an important case for examining how the edTPA acts as an individual licensure exam.That distinction has real consequences for the profession, programs, and, particularly for those learning to teach.We know relatively little about this because most of the research done on teacher performance assessments has been conducted in a different context where the stakes were lower for individual candidates and local evaluators retained some control over judgments of quality.Recent research is now revealing the impacts of higher stakes in the administration of a nationalized, commercialized teacher performance assessment.Such research suggests that performance assessments can shape candidates' experiences of learning to teach through changes to the teacher education curriculum (Ledwell & Oyler, 2016) and the nature of student teaching experience itself (Clayton, 2015;Meuwissen & Choppin, 2015).
This study explores how perceptions of teaching candidates evolved over time as, presumably, the edTPA became settled into program structures and routines over the first three semesters of its implementation as a licensure exam.The findings reveal that, while candidates' perceptions of program alignment and benefits of the edTPA improved over time, a perceived subtractive experience of the edTPA endured over this period.After a review of conceptual and research literature, the methods and results will be presented and a discussion of implications for policy and practice will follow.

An Analytical Framework: Learning to Teach for Performance
The notion of subtractive schooling (Valenzuela, 2010) stems from literature that describes the narrowed curricular and pedagogical experiences resulting from high-stakes standardized exams in urban schooling.In particular, Valenzuela described how these practices, devoid of cultural relevance, actually created schooling experiences that strip away consideration of student identities as a part of learning.One could assume that subtractive schooling is the result of defensive teaching which McNeil (1982) described as teachers' efforts to narrow curricular experiences in order to secure better standardized test scores with the knowledge that such practices do not improve deep understanding.Given these observations of student learning, what happens to teacher learning in such environments?
In the context of educational reform that produces subtractive schooling and defensive teaching, Ball's (2003) discussion of teacher performativity provides a lens for the conceptualization of this study.Ball discussed performativity as "a technology, a culture and a mode of regulation that employs judgments, comparisons and displays" (p.216).Performativity describes the orientation of neoliberal education reforms where "rewards and sanctions" (p.216) for performance act as a key tool for leverage.Ball asserted that the implications for the profession are profound where who controls the "field of judgment" (p.216) is contested and the nature of who the teacher ischaracterized by the struggle for the "teacher's soul" -is at stake.What are the consequences of performativity, hyper-rationalized by edTPA licensure policies, for candidates learning to teach and for the experience of student teaching itself?
This inquiry is also driven by ideas of learning to teach grounded in constructivist understandings about the fragility and uncertainty of that process (Clandinin & Connelly, 1992;Cochran-Smith & Lytle, 1999, 2009;Feiman-Nemser, 2001).Learning to teach is much more than a technical task.It is more about "uncertainty than certainty, more about posing problems and dilemmas than with solving them" (Cochran-Smith & Lytle, 1999, p. 294) and more about "forming and re-forming frameworks for understanding practice" (pp. 290-291).Moreover, the learning to teach process is complex, requiring a simultaneous development and integration of content and pedagogy, theory and practice.As they learn to teach, candidates need to shift towards a teacher perspective and develop the capacity for complex decision-making in action, taking into consideration a variety of student needs and content considerations (Darling-Hammond, Hammerness, Grossman, Rust, & Shulman, 2005).
Together, these perspectives on accountability, teacher performativity, and learning to teach inform assumptions undergirding the study.These perspectives prioritize investigation of the perceptions of those living through these experiences as they offer insight into what candidates are actually learning, who they are actually becoming, and how the student teaching experience is evolving.

Research on Teacher Performance Assessment and Student Teaching
Grounded in this theoretical landscape, the following literature review considers research on teacher performance assessment and particularly its interaction during the critical moment of student teaching.Teacher education programs have long confronted the challenge of translating theory from coursework into applied practices with students and communities in schools.Student teaching is the moment in teacher education when these two worlds collide (Feiman-Nemser & Buchmann, 1983).Student teaching is a critical component of learning to teach (Borko & Mayfield, 1995;Zeichner, 2002) and provides an "experiential base" for university course concepts, theories, and practices to be examined and tested by candidates (Clift & Brady, 2005, p. 315).This direct experience shapes the beliefs and attitudes of pre-service teachers in both progressive and regressive ways (Anderson & Stillman, 2011).
The quality of the student teaching placement has often been determined by the quality of the relationships between university supervisor, the mentor teacher, and the candidate (Zeichner, 2002).Research documents many tensions that plague these relationships and, thus, student teaching.Such tensions include consensus about the purpose of mentoring (Slick, 1997) and the development of shared goals and trust (Bullough & Draper, 2004;Griffin, 1989;Hoy & Woolfolk, 1989).Most of this research focuses on perspectives of mentors and university supervisors (Clift & Brady, 2005), muting the voices of prospective teachers.
Within the context of this unique moment in teacher development and these fragile relationships, teacher performance assessments have been introduced.Their development has its origins in both teacher education and teacher evaluation policies and practices.This origin makes the placement of teacher performance assessments in student teaching potentially conflicted and confounding.After a discussion of the development and uses of teacher performance assessment, this review will present initial and more recent research on the edTPA, specifically, in order to suggest that promised benefits of the assessment are not fully realized in current large-scale and commercialized implementation of the edTPA as an individual licensure exam for initial certification.

Development and Uses of Teacher Performance Assessments
Teacher performance assessments have been developed for a variety of purposes and participants.While the research suggests general learning benefits for individuals and programs, a shift in recent uses for pre-service teachers during student teaching flag some concerns.
Within pre-service education, portfolio assessments have a long history of use by education faculty to document candidate growth over time and promote program changes.Faculty proponents argued that such portfolios more adequately assess the complexity of teaching, in general, and learning to teach, particularly (Wei & Pecheone, 2010;Wolf & Dietz, 1998).Critics raised questions about whether portfolios measure teaching skill or writing and presentations about teaching (Meeus, Petegem, & Engles, 2009).In an attempt to mitigate that criticism, videos have been introduced as additional components, only creating new dilemmas as teaching performance is confined to the limits of what the lens can capture.
Among in-service teachers, both the Beginning Educator Support and Training Program [BEST] in Connecticut and the National Board process for advanced certification of experienced teachers demonstrate more formalized state and national approaches to teacher performance assessment.Of these two, research on the National Board has been touted as a justification for edTPA design and use.Consisting of a structured portfolio with video, an analysis of student work, reflective commentary, and a content knowledge test, the National Board performance assessment has raised similar concerns, from some scholars, about whether the extensive portfolio measures expert teaching or the ability to write well about teaching (Burroughs, Schwarts, & Hendricks-Lee;2000;Van Driel, Beijaard, & Verloop, 2001).Limited research and commentary on the effect of Board certified teachers on student learning has generally shown enhanced teacher development (Sato, Hyler, & Monte-Sano, 2002;Whitman, 2002) and effects on student test scores (Cavalluzzo, 2004;Goldhaber, Perry, & Anthony, 2004;Vandervoort, Amrein-Beardlsy, & Berliner, 2004).Even cautious conclusions drawn from this research suggest that the development of a nationalized, standards-based portfolio assessment has contributed to improved practice and learning outcomes for expert teachers and their students.
In recent years, the Performance Assessment for California Teachers [PACT] has been developed as an option to fulfill a state requirement for program accreditation and completion in pre-service programs.Given the size and breadth of use in California, more extensive research has been conducted and, consequently, used to model the edTPA.Some research on PACT, conducted by those later associated with edTPA development through SCALE, suggested benefits for teacher learning and program change particularly through the development of a universal language of teaching practice quality (Darling-Hammond, Newton, & Wei, 2010;Pecheone & Chung, 2006;Peck, Singler-Gabella, Sloan, & Lind, 2014).Other research has raised questions about these generalizations.For example, some researchers note the unintended consequences impacting the scope of teacher education coursework and time for classroom and school duties during student teaching (Okhremtchouk, Newell, & Rosa, 2013).Others observed that much of the PACT research was done in situations where the stakes were low for individuals (Ledwell & Oyler, 2016) and locally assessed (Hébert, 2017), skewing whatever consequences might arise when the stakes are higher.
Some researchers have explored the complications of using PACT as a summative measure.For example, Sandholz & Shea (2012) found inconsistencies between university supervisor's predictions and candidates' actual PACT scores.Further examination of this discrepancy provided no clear explanation, leading Sandholz (2012) to raise familiar questions about whether the assessment measures writing ability or teaching performance.The author cautioned against losing "multiple perspectives" on teacher quality, typically provided from the field through supervisors and mentors (Sandholz, 2012, p. 124).Sandholz suggested that the nature of the stakes of the assessment may mediate how teaching candidates engage with that assessment and, by extension, their experience of student teaching.
In a validation study of PACT, Duckor, Castellano, Tellez, Wihardini, & Wilson (2014) indicate sufficient evidence to support PACT's "continued, but limited" (p.402) use for summative decisions such as licensure due to the bluntness of the instrument.However, their findings raised questions about the sub-scores, and particularly any use of sub-scores to provide educative support for teacher learning.Lastly, Margolis and Doring (2013) drew on student teachers' reflections to conclude that there seemed to be a "tipping point" (p.283) where the burden of summative requirements became detrimental for candidate learning.
In summary, research on teacher performance assessment suggests benefits for individuals, programs, and the profession.However, recent research that examines implementation of nationalized performance assessments raises dilemmas about the complexities of high-stakes teacher performance assessment policies in the context of learning to teach.

EdTPA Design and Implementation: Initial Views into Student Teaching
EdTPA studies are now beginning to build a portrait of how the assessment mediates student teaching.In a study of high and low performance submissions of the edTPA, Denton (2013) demonstrated a correlation between the length of written commentaries and high edTPA scores.Successful candidates utilized strategies disassociated from what they learned in coursework to manage the edTPA as a test such as maximizing page lengths and "scripted interactions" (p.32) in lessons to mirror rubric descriptions.In Ohio where candidates took the edTPA as a requirement for program completion, Coloma (2015) reported overwhelmingly negative perceptions of candidates who questioned the judgments of distant scorers as well as the nature of teaching embodied in the assessment itself.
In New York and Washington, Meuwissen and Choppin (2015) identified several kinds of tensions experienced by student teachers while taking the edTPA as a licensure exam.They concluded that, at this stage of implementation, the tensions were "not necessarily productive towards the ends of improving teaching and student learning" (p.19).Drawing from the same data set, a later article by the same researchers detailed how candidates adopt different approaches, often driven by consequential policies in those states, towards how they represent their teaching during the edTPA.The authors questioned whether the edTPA portfolio, in these policy contexts, was an authentic representations of candidate performance or "an indication of how they [candidates] interpret and respond to required performance criteria" (2017, p. 605).
Also from New York, both Greenblatt and O'Hara (2015) and Ledwell and Oyler (2016) reported several concerns as teacher educators.Greenblatt and O'Hara again suggest that the high language and technological demands skewed what the assessment actually measures.Ledwell and Oyler studied edTPA implementation within several programs across one college and concluded that the edTPA narrowed the preparation curriculum across multiple programs.
Taken together, these studies suggest that edTPA implementation is more problematic in realizing the benefits that initial studies of teacher performance assessments and subsequent marketing have promised.The consequences of teacher performativity, as rationalized through policies that nationalize and standardize the edTPA for high-stakes purposes, seem to indicate dilemmas in these circumstances that are worthy of more detailed and longitudinal examinations.

Methods
Grounded in the analytical framework described above and building on the research literature, the current study seeks to expand on a previous inquiry (Clayton, 2015) that focused on the first two rounds of implementing the edTPA as a licensure exam in New York.That study identified that individuals candidates experienced the edTPA as a reflective, educative, mandated, and/or subtractive experience.Candidates also reported that their student teaching experience was narrowed as they focused on completing a task with high-stakes consequences.These initial findings inspired the current investigation, the purpose of which is to explore how these experiences shift and persist over the first three rounds of edTPA as a licensure exam.In other words, after three rounds of implementation of edTPA as a high-stakes exam for initial certification, what remains and what has changed about how teaching candidates experienced student teaching?

Study Design and Context
The study employs mixed methodologies that are primarily situated within the case of a single School of Education with multiple initial certification programs in early childhood and special education (birth-grade 2), childhood (grades 1-6), and adolescent education (grades 7-12).Data sources included candidate score results and sample work provided by the unit, survey results of all eligible candidates, and interviews of nine candidates across the three cohorts from the two largest programs in childhood and adolescent education.
Within this context, this case study (Merriam, 1998) employed a sequential mixed-method approach (Creswell, 2014).An initial survey of all possible participants across three semesters was administered, from which willing interview candidates were solicited.
Participants in the three programs also consisted of undergraduate and graduate students.While the pre-student teaching fieldwork is more substantial for the undergraduates than the graduate students, the student teaching placement of 14 weeks requires two placements in the certification area and is supervised by a school-based mentor and university supervisor for both populations.For graduate candidates in particular, pre-student teaching fieldwork is loosely organized and often consists primarily of observations; thus, student teaching is potentially the first time they are performing significant teaching tasks with relative autonomy.It is also when both groups are completing their edTPA for state licensure.
The author is a faculty member in the School of Education who possessed knowledge of program goals and edTPA implementation during this period but had no duties related to student teaching or edTPA implementation.The author had some of the candidates in this study as students in program coursework; however, that interaction never directly involved work with the edTPA.This author's positionality impacted the study's design because of familiarity with program coursework, the edTPA, and persistent institutional concerns.The author's position as a teacher educator likely impacted the interpretation of results within the larger picture of teacher education accountability reform dilemmas.The author checked these interpretations to ensure they were grounded in the data through multiple presentations within the School of Education and the broader teacher education community.

Data Sources, Instrument and Protocol Design
Both quantitative and qualitative data were collected over three semesters.Data sources included candidate survey responses and institutional score results.Interviews of a smaller sample were a principal source of this investigation along with sample work related to those interviewed candidates.
First, an electronic survey was created and administered to all eligible candidates across three cohorts.The survey consisted of 50 selected-response items that included both 3-and 4-point Likert scale questions as well as two open-ended prompts.Survey items were developed based on initial literature research and three focus group discussions with candidates from the first cohort before the end of the first semester of implementation.As a result, emergent themes were mapped to identify three content categories for Likert scale questions: the nature of the student teaching placement (3 questions), knowledge and experience with the edTPA (21 questions), and relationships with mentor teachers and students (19 questions).Seven additional selected-response questions solicited demographic information and participation in a follow-up interview.In addition, two open-ended questions about impact and perception in relation to receipt of resulting scores rounded out the survey.
A draft survey was reviewed by two colleagues with experience in evaluation, assessment, and teacher education.One was a faculty member who was familiar with edTPA, program curriculum, fieldwork, and surveys.The other was the assessment director who had expertise in survey design and content relating to accreditation, certification, and, specifically, the edTPA.The draft survey was also piloted with a current student who had not yet taken the edTPA solely for improving the survey language.Reviewer feedback helped to ensure both clarity and content integrity of the questions.Reviewers suggested adding some prompts to assess the role of the edTPA in helping candidates understand the profession and to assess edTPA with program alignment while eliminating others that utilized confusing or vague educational jargon and valueladen terms.Finally, reviewers helped to eliminate redundant questions to reduce survey fatigue, encouraged using a 4-point Likert scale over a 5-point scale, and influenced survey formatting to encourage consistency.They also suggested additional demographic categories than the initial identifiers proposed.In these ways, the review strengthened the validity of the survey instrument.
An interview protocol involved nine open-ended questions designed to leverage in-depth interviewing (Seidman, 2006) for further insight into items surfaced in pilot focus groups and survey as well as individual case stories.The themes of the protocol reflected survey categories but were designed to allow greater probing for elaboration of the following: the nature of their student teaching placement and edTPA experience, relationships during student teaching, discussion of their edTPA submission and scores, and, finally, reflection on personal learning from the experience.Finally, program documents such as institutional score reports, including individual results and work samples for interviewed candidates, were also reviewed.

Data Collection
First, the survey was administered over three semesters with assurances of confidentiality; however, tracking was done in order to send out three solicitations to maximize responses.Among 182 possible participants over two semesters, 82 teaching candidates responded to the survey with an overall response rate of 45% (See Table 1).1After the second administration, the survey was adjusted as five questions about candidates' perceptions of their preparation for edTPA tasks were deemed less useful in light of other questions and the score reports.The survey now had 45 selected-response items and two short answers.that some candidates did not submit edTPA's to Pearson or some had not yet received a score at the time of this analysis.Candidates have up to a year to submit the edTPA.The edTPA is a certification requirement but not a requirement of graduation.They were solicited for a survey response and interview participation.**The Adjusted Response Rate is the rate of survey response when calculated among the total edTPA completers rather than the total initial certification populations which include those for which no edTPA scores have been received.***There is one respondent in cohort three who was the first completer of a TESOL program.Though those responses were represented in the overall sample, the size was too small to report by program.
Second, all candidates were solicited to participate in follow-up interviews.After email and phone follow-up, nine candidates from the two largest programs on both campuses in childhood and adolescent education participated in an interview for this study (see Table 2).These one-hour interviews were recorded, with consent, and transcribed.The candidates' edTPA submissions were reviewed to confirm research impressions and statements made during the interview.

Data Analysis
Quantitative analysis first produced descriptions of the overall sample.Cronbach's analysis was conducted after two administrations of the survey to ensure that items were consistently interpreted by candidates.Excluding demographic and interview solicitation items, 43 selectedresponse questions included just four questions that used a 3-point scale rather than a 4-point scale.Consequently, separate Cronbach reliability analyses were run on the thirty-nine 4-point questions and four 3-point questions, yielding alpha reliability measures of .77and .39,respectively.This analysis led to the exclusion from the analysis of the four questions deemed unreliable.Moreover, an analysis of the three content categories, after this exclusion, yielded these results for alpha reliability: nature of student teaching placement (.80), knowledge and experience with the edTPA (.82), and relationships with mentor teachers and students (.79).Generally, measures above .70indicate reliability.
These data were disaggregated by semester to note patterns and outliers, particularly in relation to four previously identified themes -reflective, educative, mandated, and subtractive experiences (Clayton, 2015).These comparisons were limited because of unequal population sizes as more teaching candidates engaged in student teaching during the spring semesters (cohorts 1 and 3) than in the fall semester (cohort 2).As a result, the power of the sample itself limits possible interpretations of these quantitative data to more descriptive analyses.The qualitative data provided complimentary perspective that displayed candidate voices in a raw and revealing form that was, ultimately, of more value in interpreting results of this inquiry.
Qualitative analysis of both open-ended survey question responses and interviews were coded according to four previously identified themes -and then read to notice additional emergent codes (Glaser, 1992) particularly in relation to themes that persisted.Coding of qualitative data was followed by analytic memo writing (Lempert, 2012) and re-readings, including categorization according to cohort, to identify how the experiences of these themes persisted and/or evolved over time.Readings of institutional score reports and candidate work samples were reviewed primarily to verify and clarify the experiences of candidates as revealed in their surveys and interviews but without the same systematic coding process used with other qualitative data.Finally, the analytic framework focused attention on particular qualitative explanations offered by candidates.This led to noticing statements that evoked reflections of how candidates managed the performance experiences and perceived themselves in relation to it.

Results
In spite of some of the critiques of fast-track implementation, these programs overall demonstrate relatively strong pass (above 80%) and mastery rates (above 30%) consistent with New York state pass rates at 80% (Meuwissen & Choppin, 2017).Behind those results, however, is the story that candidates tell over time.While certain aspects of these stories were to be expected in terms of improved program alignment and implementation over time, what remains surprising is a consistent finding that the mandated experience of the edTPA subtracts from student teaching in some qualitatively distinct ways.This seems to undermine some of the other benefits of including the assessment during this particular developmental moment of learning to teach.The persistence of the edTPA as a mandated and subtractive experience, even in spite of its reflective and educative benefits, raise some potential concerns about the ways learning to teach is changing.
This section presents results organized around observations of the data about implementation, perceived benefits, perceived costs, and critiques emerging from candidates.Data is presented in aggregate or by cohort.When individual candidate data are discussed, they are presented alongside the larger data set to situate how representative the responder's sentiments are.The use of pseudonyms helps to track the qualitative contributions of interviewed candidates as well as qualitative survey responses that elaborate on aggregate data from the survey.
In viewing the voices of student teachers, the tendency might be to dismiss critiques as evidence of candidates not motivated or capable of meeting high standards.Student teachers' perceptions are presented here not to suggest the accuracy of their claims but rather to raise consideration of how these perceptions impact teacher learning as intended by the original edTPA design.

Implementation over Three Semesters
A number of indicators in the survey data demonstrated that program alignment improved and candidates felt more prepared for the edTPA over the three semesters.First, more candidates indicated awareness of the edTPA with 27% in the third semester saying they had not heard about it prior to student teaching in contrast to 60% in the first cohort (see Table 3).*Five questions were eliminated from the survey after administration with the first and second cohorts because they yielded no useful information relative to score reports.**Ten of 11 overall survey respondents replied to these questions.***Forty-one of 46 overall survey responded replied to these questions.
Second, candidates signaled that they were more prepared for student teaching and for the edTPA over time.While candidates generally felt prepared to teach students from their program experience, this increased from 76% in the first semester to 91% in the third (see Table 4).
*Forty-four of forty-six overall survey respondents replied to these questions.
More notably, 70% of these same participants felt well prepared in their programs for the edTPA, an increase from the 48% of candidates in the first semester who indicated the same at the time (Table 3).
Other indicators revealed challenges in implementation that persisted over time.First, candidate perceptions regarding their scores and alignment with prior feedback received through coursework and the field showed scant improvement.Over three semesters, an average of 58% of respondents felt that their scores were consistent with prior program feedback obtained through field supervision or coursework (Table 3).These data imply that approximately 40% of respondents felt some misalignment between their edTPA assessment and prior program-oriented assessments provided by faculty and clinical supervisors.
Second, student teachers' mentors lacked knowledge regarding the edTPA; this perception persisted even though the edTPA also remained a dominant feature of candidates' relationships with their mentors.While candidates were generally favorable (and, indeed, grateful) for their schoolbased mentors, or cooperating teachers, and attributed much of their learning to that relationship, an average of 74% of candidates consistently reported over three semesters that their mentors lacked knowledge of the edTPA (see Table 5).At the same time, an average of 73% reported that they spent some or frequent amounts of time talking about the edTPA and 63% reported that completing the edTPA became the central focus of interaction with those same mentors.This finding is interesting, raising questions about the quality of the interactions between mentors and student teachers around the edTPA.
Open-ended survey responses and interviews shed more light on this interaction.Candidates only mentioned mentors in survey comments where there were larger numbers surveyed in the first and third semesters.Among the first cohort, just five comments, or 10%, of 49 distinctly coded statements addressed mentors where the comments referenced mentors' support of their student teaching or lack of knowledge about edTPA.At least two of these comments questioned the scores referring to their mentor's assessment of their teaching in contrast to lower and failing edTPA scores received.At least two other comments mentioned that mentors did not understand edTPA expectations.In interviews, both Carol (MST, childhood education, cohort 1) and Jenny (MST, adolescent science education, cohort 2) shared about having to explain the edTPA to their mentors while also trying to complete it for certification.
In the third semester, eight comments, or about 13%, of 63 distinctly coded statements discussed mentor teachers.Many of these comments asserted that mentors were critical to their experience as student teachers and, specifically, their edTPA success.When mentors lacked knowledge about the edTPA, this left student teachers at a disadvantage and sometimes in the position, as the candidates above, of acting as a translator of the edTPA to the mentor.Nellie (MST, adolescent Visual Arts education, cohort 3) reported how she was given little guidance or clarity about her subject-specific edTPA from her mentor or the program.Half way through the first placement, her mentor asked, "What's this edTPA you keep talking about?"The student went on to report: "I shared everything I knew with her but it was like the blind leading the blind."From the first to the third semester, mentors were recognized by student teachers as a critical element of their success in student teaching, in general, and in the edTPA, in particular.Yet the persistence of mentors' lack of knowledge about the edTPA was striking, putting candidates in the position of translating the expectations about the edTPA to those who were also mentoring them.

Perceptions of edTPA Benefits over Time
Across three rounds of implementation, student teachers identified a number of benefits of participating in the edTPA during student teaching.Their comments generally identified reflective and educative benefits, which are also described by edTPA developers themselves (SCALE, 2015b).In particular, a majority of candidates consistently noted over time that completing the edTPA helped refine their understanding of their own development as educators, professional roles and responsibilities, and how content is taught in the classroom (Table 3).Candidate comments on the survey were also confirmed in interviews and revealed how the edTPA was a "tool for reflection" and that sometimes a low score -even a failing one -caused the candidate to reflect on an aspect of practice where they needed to work more.
Candidates also cited some of the educative benefits of the edTPA, centered on learning more about how to meet student needs.In a previous study, candidates from the first two cohorts generally acknowledged learning about differentiation and assessment in university coursework; however, their implementation in practice while completing their edTPA helped them realize a deeper understanding and purpose for these practices in meeting diverse student needs (Clayton, 2015).The edTPA has a number of structural elements in the context for learning, planning, and assessment tasks that require concentration on meeting students' varied needs, differentiated planning and formative and summative assessment for instruction A large majority of open-ended survey responses that were coded as educative confirmed the earlier finding (See Table 6).Over three semesters, about 30% of positive comments identified by candidates cited educative benefits.Of these educative benefits, 78% of educative codes identified differentiation and/or assessment as what they learned from completing the edTPA.For example, a third semester candidate noted, "The edTPA did help me to focus on my students with special needs, and was a reminder that I had to incorporate more positive reinforcement in the classroom."Karen (MST, childhood education, cohort 2) failed the edTPA but still reflected, during her interview, on how her failed score taught her to pay increased attention to assessment and differentiation.Karen also noted, on her survey, that completing the edTPA enhanced her student teaching experience because "I have seen the fruition of what happens when you do modify your lessons to accommodate different groups of individuals."Both Jenny (MST, adolescent science education, cohort 2) and Kristina (BA, adolescent math education, cohort 3) spoke about how much they learned as they looked back on the edTPA experience but not particularly while they were in it.They both spoke, specifically, about how they learned much about assessment and differentiated instruction but that expectations for these practices were more sophisticated than what their mentors were modelling in their high school classrooms.
Understanding student needs was an educative benefit often linked to candidates' interests in differentiation and assessment.While 61% of respondents in the third cohort said that completing the edTPA helped them understand students' needs and interests, up from 48% in the first cohort (Table 5), these particular benefits did not generally persist in either the survey data or interviews by the third semester.For example, most survey indicators that identified how the edTPA helped students develop understandings about students' needs, learning, and the context for learning declined over the three cohorts (Table 3).In the first semester, 64% of respondents said that the edTPA helped them understand student needs and 71% reported that the edTPA helped them better understand the context for learning.By the third semester, just 39% said the edTPA helped them understand student needs and 28% said it helped them understand how the context contributes to learning (Table 3).
Across three semesters of the quantitative survey data, candidates generally reported that the edTPA did not particularly afford them more opportunities to teach (Table 4).Open-ended survey responses, however, revealed that several individual candidates experienced qualitative and quantitative differences in teaching opportunities as a result of the edTPA by the third semester.Of 12 statements coded as positive benefits of the edTPA during cohort 3, the largest category -with five comments -were candidates who discussed how the edTPA afforded more opportunities for teaching (Table 6).For example, one candidate spoke about developing "more responsibilities" while another spoke of "added opportunities" in specific components of the tasks.Others reported about having "full control" and planning and assessing their own lesson plans, rather than delivering and assessing their mentors' lesson plans.
These comments contrasted sharply with Jim (MST, adolescent science education, cohort 2) who, during his interview, shared experiences of having to work within the mentor teacher's curricular approach to fit in some time to fulfill a lesson that would require perceived expectations for the edTPA.Others, in the survey, over the first two semesters discussed having to follow the mentor's pacing and curricular materials that impacted their edTPA submissions.While these data suggest some shifts, no clear explanations for why this change occurred by the third semester is evident.

Persistence of a Mandated Experience
Across three semesters, teaching candidates were generally positive about their mentors, their experiences with students, and their program preparation for student teaching as an average of 86% reported they felt prepared to teach by their programs (Table 4).While they spoke about its benefits, candidates also cast those reflections within a view of the edTPA as a requirement that needed to be fulfilled.These revelations were most evident in written survey comments or interviews.
Among the open-ended survey responses to one question, 20 comments, or 45%, were coded as discussing the edTPA as a mandate in negative terms (Table 6).Candidates spoke about being too busy "worrying about completing templates," video editing, and "endless write ups (sic)."Many discussed the edTPA in opposition to student teaching.One candidate from the first cohort said, "The entire experience of student teaching was too much about writing the edTPA and it really hurt the entire experience of teaching."Others went further and said it "interfered."As one candidate in the third semester concluded, "Completing the edTPA led me to a student teaching experience focused largely on completing the edTPA." In these survey responses, candidates specifically addressed the way their planning was altered due to the edTPA.One candidate in the second semester wrote that completing the edTPA This statement suggests how the nature of the requirement impacted candidate choices and to some degree how the stakes of this performance seemed to alter what candidates did -even if the candidate did not believe it necessarily benefitted students.Finally, a candidate from the first semester summed up the impact: "I was so focused on completing the edTPA and meeting all of the requirements that as a result, I was not able to plan lessons how I would have originally done if I was not completing the edTPA."During interviews, candidates across the cohorts further elaborated on how their desire to complete the requirement led them to focus on certain elements and neglect others in their disciplines.Childhood education candidates, in particular, acknowledged focusing more on math and literacy while neglecting other content areas not the focus of edTPA tasks.For example, Carol (MST, childhood education, cohort 1) spoke about focusing her student teaching on these subjects since they were the focus of the elementary edTPA.Similarly, this concern was reflected in other program areas when Jim (MST, adolescent science education, cohort 2) shared that he found himself "trying to shoehorn in…instruction that would meet the requirements of the prompts in the edTPA into an existing curriculum."Greg (MST, adolescent science education cohort 3) suggested that the edTPA prompts were more "slanted towards a humanities based classroom" that valued interpretation over the analytical skills of his discipline.More bluntly, Kristina (BA, adolescent math education, cohort 3) shared some of the questions she asked herself as she prepared her lessons: "What do I need to do to pass?Did I remember to do …the requirements that they needed for the math section?Am I giving the State what they want or am I actually teaching the way a real teacher is going to teach in the future?" To a certain degree, the experience of the edTPA as a mandate is not surprising; it is a requirement so the prevalence of that theme reported by candidates would be expected.From the start, however, candidates discussed the requirement as taking away from some perception of what their student teaching experience should be.Candidates suggested that the edTPA took away from the "practice" of teaching; others asserted it left less time for learning through reflection because they were "so caught up in this test."While there was some indication in survey responses of increased opportunity and autonomy during student teaching by the third semester, many others had earlier shared how the edTPA interrupted daily student teaching.For example, Carol (MST, childhood education, cohort 1) shared that her mentor teacher allowed her to teach less so she could sit in the back of the room and write her edTPA.Her focus was inward -on herself -and, with students, she worried about what she could get out of the kids that would benefit her submission.This changed once the edTPA was behind her and she could be more fully present with the elementary students with whom she worked.

When Mastery is not Enough: Subtracting from Learning to Teach
This sense of the edTPA as a subtractive experience persisted and appeared to strengthen by the third semester where 50% of negative comments for one survey question were coded subtractive, an increase in proportion from previous semesters (Table 6).Among representative comments, candidates talked about being "consumed" and "not able to gain as much as" they felt they should have.One candidate, while acknowledging the meaningful intent of the edTPA process and its role in ensuring that he/she focused on students with special needs, suggested that it "robbed" the candidate of some expected experience where he/she "could not fully be myself in the classroom" because of what needed to be included on the video.Several spoke about being "scared of failing" or too fearful "to try new strategies and take risks."One candidate wrote that the edTPA "limited my ability to try new things in the classroom as I was focused on the edTPA performance."In this sense, the candidate's use of the word "performance" reveals the candidate's preoccupation with passing the exam rather than as a tool to improve teaching practice.These perceptions seemed to be represented similarly across programs, cohorts, and score achievement levels.
Across semesters, candidates discussed the edTPA as taking something qualitatively away from their student teaching experiences and, in interviews, some tried to articulate more what that was.Mick (BA, adolescent social studies education, cohort 1) discussed that he spent hours working on choosing the right words in his reflective commentaries; this led to feelings of frustration about time lost for other things: "I could be using that [time] to read about new methods of teaching, possibly implement them, and find -not find -but to create new lessons."Upon receiving a nearperfect mastery score, Mick shared a sense of disbelief.He worked closely on his submission with a roommate who happened to fail.He admitted that his success made him "doubt myself a little bit" as if he got an easy scorer or it wasn't as hard as he had expected.The entire experience made Mick wonder if the time spent worrying about the language of his edTPA submission was so necessary and, indeed, helpful to his growth as a student teacher.
Coryn (BA, childhood education, cohort 3), who also received a mastery score, acknowledged that the edTPA could be "rewarding" and was a fairer approximation of what it takes to teach, in contrast to existing certification exams.She also spoke in exasperated tones about the edTPA.Echoing others' comments in the survey, Coryn explained that her work on the edTPA placed so much stress during an already critical moment of learning that it "took away from student teaching."When pressed to elaborate, she shared that she was translating the methods she had learned for the first time in practice with real students in a consistent and daily way.She suggested that a focus on the edTPA in this moment took away from that important act of translation.Combined with the lack of any feedback other than a number score, the receipt of such a score was both a source of surprise and confusion, just as it was with Mick.Though proud of their mastery scores, both felt uncertain about their successes and what they had to give up to experience the edTPA during student teaching.

Student Teachers Questioning the Exam
Over the three semesters, candidates critiqued the experience of the edTPA raising critical questions that highlighted issues of urban education and social justice as well as the meaning of the assessment.On the open-ended survey responses for one question, the proportion of negative comments coded as critique increased to 42% in the third semester from 33% in the first (Table 6).
Candidates in interviews most readily clarified their critiques.Mary (MST, adolescent science education, cohort 1) worried about how student teachers working "difficult classroom environments where we need good teachers most" might be unfairly judged after her experience of student misbehavior captured on her video clip.Mary decided to submit the clip anyway because she thought it represented her teaching and classroom accurately; she went on to narrate how she handled the comment after the video revealed something to her that she was not aware of in the moment.The rating of a "1" on classroom environment was particularly devastating as it was inconsistent with the feedback Mary had received from her field supervisor and mentors.Upon completion of her student teaching, the school was so impressed with her teaching that the principal rearranged positions to hire her.Nevertheless, the experience made Mary wonder whether there could have been a "more comprehensive picture of me" presented in her edTPA."I don't know if that really demonstrated everything I really wanted to demonstrate.Did my writing really show them who I am as a teacher?I don't know." Additionally, a similar story was shared by Jim (MST, adolescent science education, cohort 2) who worked in an urban environment.He questioned the distant relationship that potential graders had from his particular setting.He said that his program provided "an education on being an urban teacher" by virtue of its location in a large urban center and he shared that he received positive feedback from students, mentors, and the university supervisor.He questioned whether this national assessment could appropriately assess that preparation: And so for me to produce an edTPA and throw it out into the realm of 34 other states that do it and have it land in a desk somewhere else on someone that just can't make the paradigm shift to realize what's it like to be an urban educator in this city.It's fabulous.It's not any less but it's just different… I just feel like to receive a failing grade just shows that they just don't understand what I was doing in the classroom.
His comment critiqued the circumstances surrounding administration of a national, standardized performance assessment where evaluation is removed from local surroundings and transferred to distant sites with the assumption that what the assessment measures is valid and can transcend local particularities.When he failed the edTPA, his university supervisor and mentor teacher were surprised; at the time of the interview, he continued to volunteer in the school's Advanced Placement science courses as he registered to take the safety net certification test and applied for jobs.By the third cohort, these kinds of claims intensified in both surveys and interviews.Single stories were not as prevalent as were claims made by candidates that suggested some lack of trust in what the edTPA revealed about them as teachers.Candidates suggested the scoring revealed "some wiggle room for interpretation in the rubrics" while others questioned the clarity of expectations in the edTPA wondering that "some students may fail the edTPA who are effective teachers but are not able to properly delegate the time between continuing their responsibilities towards their students while also properly completing the exam." Whether any of these critiques is "right" is not the point.What is more important is what they reveal about candidates' experience of learning to teach in this context.Where performance is high-stakes, these comments suggest something else is happening to discourage the kind of stretching that novices must do to learn to teach.In spite of whether they passed or failed and in contrast to local assessments and affirmations of their teaching, candidates raised important questioned regarding the assessment's meaning for themselves and for future student teachers in diverse contexts.

Discussion
As would be expected, some things changes and others stayed the same.The study demonstrated that candidates' perceptions about program alignment with the edTPA improved.
Candidates felt they were more prepared for its expectations while, at the same time, their perceptions of how well the edTPA measured their performance and teaching quality, more generally, led to more questions.Over three semesters, the disconnect felt by candidates between local assessments of their teaching quality by supervisors and mentors and the more distant judgments represented in their edTPA scores remained a source of confusion and tension.Additionally, the lack of mentors' knowledge about edTPA persisted over time, likely contributing somewhat to that disconnect.
Finally, the study confirmed previous findings that suggested teaching candidates continued to experience the edTPA in ways they perceived narrowed the scope of their learning (Clayton, 2015).While candidates continued to experience the edTPA as a reflective, educative, and mandated experience with only slight shifts across cohorts, the persistence and, indeed, intensification of a subtractive experience of the edTPA was particularly noteworthy after other data suggested that program alignment had improved.Considering these results in light of the context of teacher performativity and learning to teach, the following sections present discussion in three critical areas.

Reifying Quality: What are Student Teachers Learning about Quality Teaching?
As in other studies (Denton, 2013;Meuwissen & Choppin, 2015, 2017), these candidates share that they emphasized certain aspects of their teaching, even particular subjects or styles of teaching, based on edTPA task prompts and requirements.As these candidates reviewed rubric descriptors carefully in order to craft language that reflected those ideas in their written commentaries, assumptions about quality implied in the tasks, question prompts, and rubrics were taken for granted, potentially reifying what quality is in teaching for those entering the profession.
Writing before high-stakes edTPA licensure policies about the impacts of subtractive schooling on teacher education, Gainer & Larrotta (2010) warned that teacher education can advance hegemonic assumptions about teaching and learning that "normalize Whiteness" (p.42) and contribute to reproducing inequalities in education.These study's results, then, call us to wonder how the high-stakes of the edTPA normalizes what counts as quality and what constitutes appropriate learning for teachers at this critical developmental moment where uncertainty (Cocharn-Smith & Lytle, 1999) is a key characteristic.In supportive cultures, uncertainty is a necessary condition for learning and novices can thrive and construct new and personalized understanding of what quality means in their classrooms.In performative systems, where the risks of uncertainty can result in not attaining a teaching license, the certainty represented in nationally sanctioned rubrics of teaching quality gets reified as the norm.Uncertainty in these circumstances is dangerous and something to be managed.

Managing Performance at All Costs: How is Student Teaching Evolving?
These data also illustrated how several candidates dealt with the consequences of managing performance while learning to teach where the result is to take away time and focus from other critical tasks of both teaching and learning to teach.Candidates discussed being more focused on their needs -for the right video angle, for good student data to make a point about their teaching, or for the right words to describe their reflections.Ball (2003) notes how performance cultures award investment of time and energy in "tactical improvements" (p.223) that enhance performance in the short term.He draws on Lyotard's (1984) work that suggests that managing experience in performance systems involves increasing time and energy on accounting in order to prove performance relative to the actual work itself.Since sanctions are so severe -in this case, denial of certification after the expense of time and money to complete a university program -the desire to manage one's performance is intensified.This is evident throughout the data where candidates, across cohorts, discuss how they spend their time and focus their planning poring over word choice and video clips sometimes to the neglect of focusing on improving their teaching immediately and attending to other needs, such as those of their students, in their placements.Indeed, one student teacher talked about how her mentor allowed her to sit in the back of the room and type her edTPA, losing valuable time to interact with students, watch her mentor, and hone her craft.In developing a submission that makes the grade for the edTPA, the shift to performance at all costs seems to be changing the experience of student teaching in subtle, but important, ways.

Values Schizophrenia: Who Are Student Teachers Becoming?
The data also illustrated that candidates were increasingly uncertain and even critical about what the distant judgment of the edTPA score meant in comparisons to judgments that are local and more contextualized.Ball (2003) writes about values schizophrenia (p.221) where teachers can participate in "game-playing, or cynical compliance" (p.222) as a way to manage the stress of their experience.These data suggest that such behavior persisted over three cohorts of candidates and regardless of program or their final edTPA scores.
Because such moves involve values about one's authenticity in teaching performance, such a state has the potential to engender cynicism and alienation from self and the profession.This is where Ball decried concern for the "teacher's soul" with consequences for an "ethical self" (p.226).When the teacher is one who is just learning the craft, the consequences are of concern for the profession.If one learns from the start of one's career to sacrifice one's authentic self in order to play the game, as candidates in this study at all score levels admitted, then how will this impact the stance one takes towards teacher evaluation going forward?Will teacher evaluation be used to comply with an annual task or to truly improve teaching and learning?To be sure, this study does not answer these questions but it, nonetheless, raises them for future consideration.

Implications
In spite of its intended and actual reflective and educative benefits, the perception of a subtractive experience persisted and intensified regardless of program identification, cohort, or score received among these candidates in these programs.In this way, the experience of the edTPA as a high-stakes exam altered the student teaching experience for these candidates.Candidates were aware that something was being reduced in order to address the edTPA even if they could not always name it clearly.Because student teaching was a culminating experience and the evaluation of the edTPA had high-stakes consequences, these perceptions seemed magnified for candidates at this critical moment of learning to teach.
The limited scope of these data speaks most poignantly to the implications for candidates and programs.Candidates represented teaching quality to fit the parameters of the edTPA as others have noted (Meuwissen & Choppin, 2017).They reported that some sense of themselves and their capacity to learn to teach was reduced in this system.As would be expected, the programs improved their alignment with edTPA requirements.Candidates spoke to feeling better prepared over time in order to fulfill the tasks and address the language of the assessment.What remained harder to bring into alignment with distant and unknown evaluators was a shared understanding of quality among those traditionally charged with oversight for learning during student teaching -mentors and university supervisors.
This work calls on teacher educators to wonder for whom this use of the edTPA benefits and to what end does it serve?For those in this study who perceived the edTPA as a subtractive experience, how did this limit the instructional choices made and, thus, the possible learning experiences realized by teaching candidates?Did a subtractive experience result in learning that was done in defense of getting through the exam?On several occasions, candidates who were critical of the edTPA acknowledged gaining educative benefits from the tasks well after the fact from their edTPA experience but not in the moment.What are the consequences for learning to teach when it happens after the fact, in the rearview mirror?

Policy for High-Stakes Teacher Performance Assessments
While it is more difficult to draw conclusions for the profession from a study with such limited size, findings do suggest that the nature of student teaching is changing as a result of the particular way this edTPA policy was enacted in New York.To be sure, there are positive changes.Institutionalizing a common language across programs about key elements of teaching such as differentiation and assessment for instruction, as noted by the candidates here, is important.These candidates drew attention to the fact that edTPA tasks focused them on these elements more than they would have otherwise done in their student teaching.They admitted that putting these elements into practice, even though they had previously learned about them in coursework, added a new dimension to their learning about these core tasks for teaching.These are not insignificant benefits that candidates claimed they gained from the experience of the edTPA.The questions, however, remain: Could they have gotten these benefits even if the policy did not require the edTPA for licensure?Did it have to be a high-stakes licensure exam for the programs to emphasize these things or for candidates to gain this kind of learning?And, finally, were these benefits worth what appeared to be lost, or subtracted, from the student teaching experience?
These questions suggest that policy makers must manage a better balance of policy goals and impacts through more attention to how the policy impacts individual candidates who are learning to teach precisely because they are learners who are just beginning to hone their craft and develop their professional identities.Again, whether candidates' critiques of the edTPA are wholly accurate is not the point.What is critical is that their perceptions of their learning here has narrowed and that, in fact, likely impacts how much they experience this task fully as a learning experience and not merely an act of compliance.When the policy emphasizes individual candidate performance for proof of readiness for a license, the policy potentially distorts systems.In this instance, performance results at all costs, instructional moves without mistakes, and compliance over learning and program improvement can make the proposed benefits of teacher performance assessment not worth the costs.
Indeed, the field can yield the benefits of developing a profession with a shared language about teaching quality appropriate for pre-service teachers.Creating policies that require programs to implement teacher performance assessment and use the results in some substantive way to evaluate and improve programs may re-orient the policy back towards its original intent of learning for candidates and programs.Such policies could require programs to institute the edTPA, or any agreed upon teacher performance assessment.The result could be that local faculty incorporate elements into the program, determine how and where it occurs, and determine locally what kinds of consequences, if any, occur for candidates' participation.Local evaluators -that is, teacher education faculty, university supervisors and school-based mentors -together use the performance assessment to provide insight about their pre-service candidates' practice over time and guide programmatic decision making and professional development for all those involved in the preparation of pre-service candidates.This would complement current expectations for clinical partnership and use of meaningful data for program improvement demanded by the Council for Accreditation of Educator Preparation.
Mandates, whether requirements of programs or the state, always create stress for individual candidates and for programs.This study, however, begs us to ask: Would some of the subtractive experiences of this particular enactment of edTPA policy for licensure subside if the stakes were changed?Reducing the stakes for individual candidates and relocating those more limited stakes onto programs may re-orient a policy that has taken on an out-sized influence for candidates who are learning to teach.In this way, performance assessment has the potential to realize its proponents' intended goals to honor the complexity of both teaching and learning to teach.Instead of subtracting from the process, these shifts might provide more support to pre-service candidates and programs in actually learning how to become better at the process of learning to teach while negating the current policy approach's distortions of performance.

About the Guest Editors
Elena Aydarova Auburn University eza0029@auburn.eduWebsite: https://elenaaydarova.comElena Aydarova is Assistant Professor of Social Foundations at the Department of Educational "really restricted what I wanted to do or gave almost no freedom to what I really wanted to do in the classroom to explore more of my teaching."One candidate from Cohort 3 wrote, …I attempted to fit everything into the model of what theedTPA was asking for (based upon the rubrics).Because of this I sometimes removed, altered, or in somewhat (sic) changed what I had initially planned in order to fulfill the edTPA requirement.…Prettymuch in completing the edTPA I was doing what was required of me regardless if I thoughtit benefited my students and my classroom.
Foundations, Leadership, and Technology at Auburn University.Her interdisciplinary research examines the interactions between global social change and the work of teachers, teaching, and teacher education through the lens of equity and social justice.Her projects have explored teacher education reforms in Russia and the US, internationalization of education, teacher retention, as well as privatization of teacher preparation.She has recently completed a book manuscript "Teacher Education Reforms as Political Theater: Policy Dramas in Neoliberal Contexts."Throughout her career, Dr. Aydarova has taught in the United States, Ukraine, China, and the United Arab Emirates.David C. Berliner Arizona State University berliner@asu.eduDavid C. Berliner is Regents' Professor Emeritus of Education at Arizona State University.Dr. Berliner is a member of the National Academy of Education (NEA), the International Education Academy (IEA), a Fellow of the Center for Advanced Study in the Behavioral Sciences, and a past president of both the American Educational Research Association (AERA) and the Division of Educational Psychology of the American Psychological Association (APA).He is the recipient of awards for distinguished contributions from APA, AERA, and the National Education Association (NEA).Dr. Berliner is co-author (with B. J. Biddle) of the best seller The Manufactured Crisis, co-author (with Ursula Casanova) of Putting Research to Work, coauthor (with Gene Glass) of 50 Myths and Lies that Threaten America's Public Schools, and co-author (with N. L. Gage) of six editions of the textbook Educational Psychology.He is co-editor of the first Handbook of Educational Psychology and the books Talks to Teachers, and Perspectives on Instructional Time.Professor Berliner has also authored more than 200 published articles, technical reports, and book chapters.He has taught at the University of Arizona, University of Massachusetts, Teachers College and Stanford University, as well as universities in Australia, Canada, The Netherlands, Spain, and Switzerland.

Table 1
Survey Respondents in Relation to Total Population of Initial Certification Candidates

Table 2
Description of Teaching Candidates Interviewed

Table 3
Knowledge and Experience with the edTPA, 15 selected response questions*

Table 3
cont.Knowledge and Experience with the edTPA, 15 selected response questions*

Table 4
Survey Results: Nature of the Student Teaching Placement, 4 selected response questions

Table 5
Survey Results: Nature of Relationships with Mentor Teachers and Students, 19 selected responses

Table 5
cont.Survey Results: Nature of Relationships with Mentor Teachers and Students, 19 selected responses

Table 6
Coding of Open-ended Survey Responses