High-Stakes Conference: Speakers' Video Presentations

Audrey Amrein-Beardsley, Arizona State University
Associate Professor of Education
Presentation Video

Audrey Amrein-Beardsley opened the conference with a definition and brief examination of value-added models (VAMs) in the context of teacher evaluation. She illustrated how VAMs are implemented across our post-Race to the Top (RttT) environment. Based on her recent research, she raised critical questions about RttT, particularly in terms of whether VAMs work as intended. She ended with an overview of Friday's presentations and suggested ways in which the audience can engage in the conference discussion.


Spyros Konstantopoulos, Michigan State University
Associate Professor and Program Coordinator of Measurement and Quantitative Methods
Presentation Video

Spyros Konstantopoulos provided a literature review about the magnitude of teacher effects (including some research on VAMs) and whether the effects vary by different levels of education or by different subject matter. The author also discussed what is known about the stability of teacher effects over time.


Thomas L. Good, University of Arizona
Professor and Department Head of Educational Psychology
Presentation Video

Why do policy makers believe that teacher effects should be stable? Thomas Good discussed classroom complexities and whether stability can be expected of teacher effects over time. Given that some teachers are consistent in their effects on student achievement over time, he pursued information on how "more effective" teachers teach. Good discussed data limitations on explaining how teachers influence student achievement. This talk concluded with a few thoughts on why our knowledge base is so limited.


Heather C. Hill, Harvard University
Professor of Education


Corinne Herlihy, Harvard University
Project Director for the National Center for Teacher Effectiveness (NCTE)
Presentation Video

Heather Hill and Corinne Herlihy examined how state and local entities intend to ensure that teacher observation systems provide valid and reliable scores. Through a document analysis and interviews with state and local education officials, the authors explored several issues that arise in observational systems, including the overall generalizability of teacher scores, the training, certification, and reliability of observers, and specifications regarding the sampling and number of lessons observed per teacher. State and local plans for ensuring score validity and reliability were assessed in light of the planned uses of scores, and in light of the AERA/APA/NCME Standards for Educational and Psychological Testing.


David C. Berliner, Arizona State University
Regent's Professor of Education Emeritus
Presentation Video

David C. Berliner discussed how existing analytic models of teacher effectiveness only can examine a few variables simultaneously, leaving a myriad of unobserved variables interacting with each other, therefore making every school year and every class taught different from another. Potentially unobserved variables in a class (for example the percent of boys, or percent of parents with higher education degrees) interact with potentially unobserved variables in a school (for example the percent of a school that is lower class, or percent of the teachers who are new to the profession). At the same time, these two classes of variables are also interacting with potentially unobserved variables in a neighborhood (for example community safety, or rate of employment). The unobserved variables are countless and thus measures of teacher effectiveness have something in common with chaos theory, namely, that initial conditions that are unobserved, quite subtle, and even quite far from the classroom, can still determine the achievement outcomes of that classroom.

Alyson L. Lavigne, Roosevelt University
Assistant Professor of Curriculum Studies
Presentation Video

The stakes are getting higher for teachers daily as more and more states adopt hiring, firing, and tenure-granting policies based on teacher evaluation scores. These high-stakes decisions are situated in the rationale that firing ineffective (as primarily measured by observation and value-added scores) teachers will improve student achievement. However, this premise is challenged by recent findings demonstrating that all teacher turnover hurts student achievement. With teacher satisfaction at an all time low, concern about the effects of high-stakes teacher evaluation on teachers and students has become increasingly important. In this talk, Lavigne examined how high-stakes teacher evaluation decisions are changing and could change the landscape of American education.

Rick Ginsberg, University of Kansas
Dean, School of Education and Professor of Educational Leadership and Policy Studies
Presentation Video

Neal Kingston, University of Kansas
Director of the Center for Educational Testing and Evaluation & Professor of Psychology and Research in Education

Demands for accountability for teacher education are growing among policy makers as the belief of the decline of American schools lingers. The field, through its accreditation practices and requirements for meeting state guidelines for licensure and certification, is heavily ensconced in labor intensive efforts at examining and assessing current practice. As leaders demand greater use of NCLB inspired outcome measures for teacher preparation, the field is caught in a vice between meeting the expectations of leaders and media clamoring for evidence of program effectiveness and the realities of the psychometric and practical shortcomings of many current assessment tools. By exploring accreditation and the use of outcome measures in a variety of professional fields, and assessing the strengths and weaknesses of potential measurement tools being considered, this paper examined the situation facing teacher education, the iatrogenic problems that invariably result from the forced implementation of inappropriate accountability treatments, and some possible avenues for the field to consider in dealing with the reality of today's accountability-driven atmosphere.

Ronald W. Marx, University of Arizona
Dean of Education & Professor of Educational Psychology

Ronald Marx spoke about the sessions and the final panel. Conference participant voices (academics, superintendents, principles, and state level policy actors) were represented in expressed concerns, issues, and resolutions.

Closing Panelists

Presentation Video




Andrew Morrill, Arizona Education Association






Deb Duvall, Arizona School Administrators
Executive Director






Jane West, American Association of Colleges for Teacher Education
Senior Vice President





Ronald W. Marx, University of Arizona
Dean of the College of Education & Professor of Educational Psychology