BEAR Center Calendar

BEAR Events Calandar

Spring 2002

|Archive of Past Seminars|Current Seminars|

12 Feb

(Tue)

Effects of Social Units on Teachers Perceived School Climate of the Secondary School ñ Multilevel Analyses

Sang-Jin Kang, UC Berkeley and Yonsei University

School reform studies advocated that features of school climate such as administrative leadership, staff collaboration, and teacher autonomy can enhance the effectiveness of schooling. These three elements of school climate are assumed to capitalize on the commitment and expertise of a professional teaching force. Studies on school climate, however, have largely ignored the internal differentiation of secondary school organization. Also, little is known about the conditions that support the emergence of school climate in schools.

This paper studied the internal differentiation of school climate features in middle schools at Seoul, Korea by investigating the effects of individual teachers, academic departments, school subjects, and schools as the influences of social units for school restructuring. Thus, the purposes of this study were: (a) to examine the effects of the four social units on school climate perceived by teachers, (b) to examine the constraints on the development of school climate at both within- and between-school levels, (c) to contribute to the methodology of school climate studies.

This study used the data of 1,652 teachers working at 52 schools randomly selected from 355 total middle schools in Seoul, Korea. A 2-level and 3-level HLM, and crossed-multilevel modeling technique were employed for the analysis. A crossed multilevel model (CMM) treats the data in the way that teachers are nested within the departments which are defined by the cross-classification of subjects and schools. The results demonstrated the extent to which those social units affect three elements of school climate. The analyses also identified the characteristics of teachers and schools that accounted for the variation of school climate measures across the social units. Discussions are provided for policy implication of findings of internally differentiated multilevel structure of secondary school climate and of the conditions on the development of desirable school climate.

27 Feb

(Wed)

Using a Model of Learner Readiness to Study the Effects of Course Design on Classroom and Online College Student Performance

Cathleen Kennedy, UC Berkeley and College of San Mateo

Why do some students do well in online classes while other, equally qualified students don't do as well? Why do some successful classroom students find online courses more difficult? Are the indicators for success in online courses the same as those for classroom courses? Is it the remote, asynchronous communications mode or the learner-centered course structure of online courses that presents the greatest challenge to online learners?

Cathy will present her model of learner readiness and describe how she used it to answer these and other questions about how the design of online courses can affect student performance and persistence.

12 Mar

(Tue)

Preview of AERA Presentations + Panelists to share experience on AERA Conference

(Details to be announced.)

9 Apr

(Tue)

DYNAMICALLY ADAPTIVE ASSESSMENT PRACTICES

Prof Bernard Gifford, President, CEO & CIO, The Distributed Learning Workshop, & Prof of Edn, UC Berkeley

11 Apr

(Thur)

A Note on the Interpretation of the 2PLM Discrimination Parameter

Francis Tuerlinckx, Department of Statistics, Columbia University

In this paper two new interpretations of the discrimination parameter in the Two Parameter Logistic model (2PLM) for two-alternative choice items are proposed by linking the 2PLM to two stochastic models. The first interpretation is based on the fact that in the diffusion model with constant drift rate and variance and two absorbing boundaries the probability of absorption equals the 2PLM if the process is unbiased. The discrimination parameter is the distance between the two absorbing boundaries and is understood in terms of the amount of information that has to be collected before a response can be given. In a second interpretation, the 2PLM is linked to a horserace model. Under certain conditions, the discrimination parameter is a measure of independence of the information underlying the response. As an illustration, both models are estimated and their fit is compared for a particular data set.

23 Apr

(Tue)

The Impact of Item Format on Depth of Cognitive Processing: Do Multiple Choice Items Elicit the Same Types of Cognitive Engagement as Open-Ended Items?

David Pearson and Diane Hamm, UC Berkeley

In this presentation, Pearson and Hamm discuss their attempts to use think-aloud protocols to determine whether there are unique cognitive processes that can only be accessed through more open-ended assessment tasks, using tasks from NAEP reading and mathematics assessments?

7 May

(Tue)

A Psychological Theory of Goal-Directed Behavior: Disentangling Subjective and Objective Influences

Florian Kaiser, University of Technology, Eindhoven, The Netherlands

People's behaviors are jointly influenced by socio-cultural boundary conditions (objective influences such as contexts), as well as people's reasons (subjective motivation). In psychology, we do not disentangle the subjective and objective influences on people's performance rigorously. Commonly, psychologists disregard a context's objective influences--its effective significance for everyone--and consider, at most, its psychologically mediated, subjective, impact. Technically speaking, contexts are usually considered in psychology by assessing subjective concepts such as self-efficacy or perceived behavioral control. Nevertheless, contextual circumstances shape our acts even when we are not motivated to become engaged and they can influence behaviors even beyond a person's subjective acknowledgment. Otherwise, nobody would ever bump his or her head on a rafter. In our research, based on a novel approach to the measurement of behavior, we are able to discriminate between objective (e.g., socio-cultural) and subjective (i.e., motivational) behavior influences. I will present data from several cross-cultural studies conducted in Sweden, California, Switzerland, Germany, and Spain. Unanimously, our data corroborate, on the one hand, significant context-specific performance impacts (regardless of people's motivation to act). On the other hand, we also found strong and cross-culturally robust motivation (i.e., attitude) effects in the prediction of behavior.

BEAR Center
Graduate School of Education
University of California, Berkeley
Berkeley, CA 94720