| Session Title: Evaluation in the Context of High Stakes Assessments |
| Multipaper Session 702 to be held in International Ballroom B on Saturday, November 10, 9:35 AM to 10:20 AM |
| Sponsored by the Pre-K - 12 Educational Evaluation TIG |
| Chair(s): |
| M David Miller, University of Florida, dmiller@coe.ufl.edu |
| Abstract: Funding for many educational programs is often tied to an evaluation of the effectiveness of the programs. With the increased emphasis on accountability and high stakes assessments as required by the No Child Left Behind legislation and similar state initiatives, policy makers often define program effectiveness in terms of growth on the high stakes required assessments. At the same time, there is a growing concern for experimental design to measure program effectiveness which has resulted in a push for large scale experimental designs with random assignment of units and well established control groups. This presentation will focus on the use of high stakes accountability data to examine the effectiveness of education programs, particularly the issues of interpreting results of experimental studies using high stakes accountability testing. |
| Design Alternatives to Measure Effectiveness of Programs With High Stakes Assessments |
| M David Miller, University of Florida, dmiller@coe.ufl.edu |
| This paper will consider issues related to data collection when adequate control groups are not possible because high stakes assessments are the required outcome measure and all schools are working to increase their scores. Designs include measurement over time and looking at growth models; regression discontinuity designs; and experimental designs with alternative treatment options. Data are reported for the Florida Reading Initiative that has been tracking 53 schools for the last five years in a state testing program that spans more than a decade. Other examples will be discussed. |
| Interpreting High Stakes Test Data: Consequential Evidence and Multiple Stakeholders |
| Jenny Bergeron, University of Florida, jennybr@ufl.edu |
| This paper focuses on the consequential evidence for valid interpretation of high stakes test scores. Within the context of evaluation, it is important to examine the interpretation of testing data particularly in high stakes testing environments. This paper reports the effects of high stakes assessments in Florida which combines results from three studies. In the first study, principals were interviewed to determining the effects of the high stakes tests on test interpretation. In the second and third studies, teachers and students were given surveys to determining the effects on instruction and student psychological variables (e.g., feelings of control over their ability to do well on tests). |