Development and Validation of an Instrument to Evaluate Science Teachers' Assessment Beliefs and Practices

The primary purpose of this study was to develop a valid and reliable instrument to examine science teachers' assessment beliefs and practices in science classrooms. The study also investigated the relationship between teachers' beliefs and practices regarding assessment issues, the certain factors, influencing their assessment practices and teachers' feelings towards high-stakes testing. The participants of the study were 408 science teachers, teaching at middle and high school levels in the State of Florida. Data were collected through two modes of administration of the instrument as a paper-and-pencil and a web-based form. The response rate for paper-and-pencil administration was estimated as 68% whereas the response for the web administration was found to be 27%. Results from the various dimensions of validity and reliability analyses revealed that the 24 item-four-factor belief and practice measures were psychometrically sound and conceptually anchored measures of science teachers' assessment beliefs and self-reported practices. Reliability estimates for the belief measure ranged from .83 to .91 whereas alpha values for the practice measure ranged from .56 to .90. Results from the multigroup analysis supported that the instrument has the same theoretical structure across both administration groups. Therefore, future researchers may use either a paper-and-pencil or web-based format of the instrument. This study underscored a discrepancy between what teachers believe and how they act in classroom settings. The majority of teachers reported that instruction time, class size, professional development activities, availability of school funding, and state testing mandates were the primary factors, influencing their assessment routines. Many teachers drew attention to several negative impacts and consequences of the high-stakes testing both on teaching and learning. Teachers stated that both the preparation process and the results of the test created unbelievable tension both on students and teachers. Implications of the study indicated that it would be valuable to conduct alignment studies to examine whether state tests are fully aligned with the state standards and classroom assessment. Perhaps, such analyses would assist state level decision makers in reconsidering the current policies and "unintended" influences of mandated tests on classroom practices.