BEAR Events Calandar

Spring 2003

|Archive of Past Seminars|Current Seminars|

28 Jan

(Tue)

Investigations of Differential Item Functioning: Comparisons among approaches, and Extension to a Multidimensional context

Dr Insu Paek, UC Berkeley

Differential item functioning (DIF) has been one of the major topics in the measurement field for the last thirty years. In this study, a unidimensional parametric item response theory (IRT) DIF method based on multidimensional random coefficient multinomial logit (MRCML) model (Adams, Wilson, & Wang, 1997) and Mooreís (1996) work is formulated, and its performance is compared with the simultaneous bias test (SIBTEST) model (Shealy and Stout, 1993) by simulation studies. SIBTEST is a non-parametric IRT DIF model. These two approaches are compared in a variety of simulated conditions. The foci of the comparisons are the behavior of DIF estimators, Type I error and Power. For the Type I error and power investigation, (asymptotic) z-test and likelihood ratio test are used in the suggested parametric method. Lastly, Using the flexibility of the MRCML model, a multidimensional extension of the unidimensional parametric DIF modeling is attempted and its performance is investigated by a simulation study. The effect of using a unidimensional DIF model in a multidimensional test on the DIF investigation is also shown by simulation study.

11 Feb

(Tue)

Linear Regression: Modeling School Performance

Dr. Sophia Rabe-Hesketh, Institute of Psychiatry, King's College, University of London

Linear regression will be introduced by considering the problem of assessing and comparing schools' effectiveness. Starting with a t-test, proceeding to simple and multiple linear regression, I will finish with a brief introduction to multilevel modeling. Emphasis will be on interpretation rather than technical details.

25 Feb

(Tue)

From Authentic Assessment to Authentic Accountability: A View of Lateral Accountability in Education

Prof Paul LeMahieu, National Writing Project/UC Berkeley

Current approaches to educational accountability privilege hierarchical forms of thought in which the blame game races up and down systems with little evidence of improvement of performance at hand. This talk will propose and examine an alternative to current hierarchical models of educational accountability, one in which accountability is approached as a constructivist undertaking in its own right. The approach values lateral forms of accountability, in which each element of the system is directly accountable to the various publics that it serves. Relevant research on human motivation will be brought to bear to inform an accountability policy and approach that genuinely seeks to motivate increased effort and improved performance. In addition the conditions (e.g., intellectual, dispositional, structural, etc.) necessary to engender this approach to accountability will be explored. Finally, a survey of extensive research on and experience with accountability systems will suggest defining issues that shape the debate about accountability policy and practice.

11 Mar

(Tue)

1-3pm

The Diversity in Math Education Project Encounters the Real World

Prof Alan H. Schoenfeld, UC Berkeley

With the arrival of the new superintendent of the Berkeley Unified School District (BUSD), Michele Lawrence, a simple plan for having a "teacher in residence" at UCB expanded into a major collaboration between BUSD and the "Diversity in Mathematics Education" (DiME) center at UCB. There is now an ongoing seminar with middle school teachers, focusing on issues of diversity and middle school mathematics. DiME is grappling with these issues, among others:

*Whether and how to evaluate the piloting of curricular materials, where teacher assignments aren't random and the curricula aren't equivalent;

* How to capture what students learn, in fine-grained detail, as part of "lesson study";

* What data to gather to document in order to capture equity-related phenomena, and chart whether the district is making progress;

* How to measure the impact of professional development.

Life isn't easy in the real world. We will describe how the project evolved and some of the issues involved in trying to make a difference through careful research in settings that won't sit still.

1 Apr

(Tue)

Topic to be Announced

Prof Kathleen Metz, UC Berkeley

15 Apr

(Tue)

Preview of AERA Presentations

29 Apr

(Tue)

Prof George A. Marcoulides -California State University, Fullerton

Title to be announced.