Design

This quasi-experimental, mixed-methods, longitudinal study (Shadish, Cook & Campbell, 2002), incorporates implementation with research, and uses quantitative and qualitative methods blended in a design research framework (Cobb, et al., 2003). TI group and comparison group teacher and student measures are collected the year before teachers enter the TI program (baseline), each year as teachers progress through TI, and post-TI. The "control" is provided by the comparison group who will be matched with the TI group on the bases of a number of individual characteristics and school demographics. As teachers have different students each year, the design is strengthened by pre- and post-testing students annually to account for the differences among incoming students from year to year. The features of this study are consistent with those described in a recent synthesis of purposes, settings, and distinguishing features of design studies (Cobb, et al., 2003). Because this study is framed to test the further development and implementation of the TI model in two derivative forms, it will operate in on a small enough scale and with consistent feedback from PD designers, teacher participants, and evaluators to identify design challenges and make improvements to the program as it is being delivered. The design research approach allows the project team to work in the PD delivery context while using focused data and formative evaluator feedback to systematically optimize each phase of each site's implementation. Although the study's plan involves one round of implementation at each site, the project, like other design studies, takes on a recursive feel due to the PI's experience with the TI model in its first two iterations in the pilot phase. Lastly, the specific theoretical orientation of this study, its targeted hypotheses, and particular implementation conditions can generate practical and domain-specific theories with respect to effective inquiry PD in science. As such, the scope of this project is consistent with design research according to Cobb and coworkers (2003) who state that design experiments are situated between a narrow theoretical scope (e.g., a particular classroom) and a general philosophical orientation to teaching and learning (e.g. constructivism).

Settings: The components of the TI model summarized in Table 2 and piloted at GVSU will be translated into two sets of programs and courses for this study: (1) GVSU - TI in biology, physics and geology in addition to chemistry; and (2) Miami - TI in chemistry. A detailed account of how each site has integrated the TI model into their graduate coursework, time line for the implementation of each component, and mechanism for TI delivery are described in detail below.

Cobb, P., Confrey, J., di Sessa, A., Lehrer, R., & Schauble, L. (2003). Design experiments in educational research. Educational Researcher, 32, 9-13.

Shadish, W., Cook, T., & Campbell, D. (2002). Experimental and quasi-experimental designs for generalized causal inference. Boston: Houghton-Mifflin Company.