Conference Agenda

Focus of Inquiry and Connection to Conference Theme

Connecting to the ICSEI 2024 theme of ‘quality professional education,’ this paper explores how science teacher learning is shaped by distinctive features of a science museum, including practicum experiences in galleries, interning in youth programs, courses co-taught by museum scientists; and doing research alongside museum scientists. This paper responds to the ICSEI sub-theme “policy and practice learning to support teacher and school leader development,” by shedding light on features of teacher education programs that may be particularly effective in preparing teachers for equitable science teaching in public schools.

Theoretical/Conceptual Perspectives

We draw upon two theoretical frameworks in the design of our teacher education program and our research within it (Authors, 2022). A sociocultural theoretical framework (Vygotsky, 1978) reflects our view of teacher learning as an interactive process of participation in a science community (Lave & Wenger, 1991); and a critical theoretical framework asks that we acknowledge that science is not neutral; rejecting deficit views; and incorporating a systemic view of inequality and injustices (Milner, 2010; 2021).

Data/Method

Data are drawn from a case study of a 15-month teacher education program that prepares Earth science secondary teachers in a large, urban intensive city, housed in a science museum. Data include interviews with program students, observation of program meetings and courses, as well as program documents. Additional data from ongoing evaluations, and findings from published research by program scientists and educators, supplements the case. Using these data, we explore what residents learn in order to enact equitable science teaching. We identify features of the setting of a science institution that play an especially important role in strengthening teachers’ science instruction and equitable work with youth.

Findings

The case study points to four distinctive features of learning in a museum that support teachers’ learning to teach science:1) an eight week “science practicum” involving preservice teachers in conducting research with active scientists; 2) early teaching experiences in museum galleries; 3) fieldwork in museum-based afterschool youth programs and 4) a co-teaching model in which science courses are taught by educators and museum scientists. These contributed to teachers’ heightened ‘science identity’; stronger grasp of science practices for geology; a heightened appreciation of and knowledge about local Earth science phenomena; and asset-based views of youth. For example, the science practicum provided science resources they could use in their classrooms, strengthened knowledge about how scientists do their work and helped counter views of science as ‘not neutral’, and strengthened their ability to communicate with students about complex science concepts.

Educational Importance for Practice and Theory

This case study reveals how a program’s features can support preservice teachers to learn ambitious and equitable science teaching. While few teacher education programs take place in a museum or a cultural institution that can support subject-specific learning; this study points to ways that teacher education programs can take advantage of informal science institutions and their resources. This study shares principles that other teacher education programs can draw upon to inform their work preparing teachers.

Problem statement

Value-added models (VAM) are widely used around the world to monitor school effects of schools’ effectiveness (Leckie & Prior, 2022). In multilevel VAM, students’ current achievement is regressed on (at least) prior achievement, the school effectiveness estimate, i.e., the value-added (VA) estimate, being the school-level residual. However, measurement error on prior achievement can lead to attenuation bias and to biased schools’ VA-estimates: the VA-estimates of high prior-achieving schools are overestimated while the effectiveness of school serving low prior achievers is underestimated (Kane, 2017; Perry, 2019).

This study investigates how ability estimators of prior achievement can affect the school-level VA-estimates. Three VAM are studied: simple value-added (VA) models, and contextualized VAM controlling for students’ sociodemographic features (CVA-A) and also for schools' prior achievement (CVA-B) (Leckie & Prior, 2022).

The investigated ability estimators rely on Item Response Theory (IRT) measurement models: weighted-likelihood estimates (WLE), expected a posteriori (EAP) and plausible values (PV). These estimators have different properties. For instance, WLE overestimate ability’s variance while EAP underestimate this variance (Lechner et al., 2021; Monseur & Adams, 2009; Wu, 2005); in multilevel models, the estimated variance components can also be affected by the estimator choice (Monseur & Adams, 2009). Furthermore, Schofield et al. (2015) showed that, when using PV as an independent variable in a regression model, biased estimates can arise if the conditioning model used to procure PV is not compatible with this regression analysis model. In the continuity of these findings, this study investigates how these ability estimators for prior achievement can affect schools’ estimated effect.

Methodology

A simulation study is conducted to investigate the influence of several prior achievement estimation methods on school residuals. The ability estimators are WLE, PV and EAP. The latter two rely on Bayesian approaches and are estimated without conditioning or conditioned on contrast coding for schools or on schools mean prior achievement (in order to account for school differences), with or without other conditioning variables.

Multilevel VA, CVA-A and CVA-B models are investigated: the bias in the correlations between the school-level residuals in each prior-ability estimate condition and (a) schools’ mean prior achievement and (b) schools’ true residuals, is analyzed. Attention is also paid to the estimated model parameters. Several test lengths for assessing prior achievement are explored; true values of current achievement and of the covariate (for CVA models) are used.

Results

Results show that using an improper ability estimator of prior achievement can bias the correlation between school prior achievement and school residuals in VA and CVA-A models. Bayesian ability approaches with conditioning variables can recover this correlation and the specification of the IRT conditioning model is discussed. This correlation is also less affected by the ability estimator choice in CVA-B models than in CVA-A and VA models, as they control for school mean prior achievement.

Inadequate prior achievement estimates can therefore lead to misleading school effects. Researchers’ choice should rely on the specification of their VAM and on the properties of ability estimators.

This paper reports on the on-going research project called Professional Identity for Khmer Teachers (PIKT). In all education settings the challenge of enhancing teaching capability to achieve better learning outcomes is complex and multifaceted. In the context of early grade education in Cambodian schools, this is especially the case. Cambodian education is like the ‘perfect storm’ of challenges.

The authors are part of the PIKT research collaboration seeking to find new ways to improve the quality of teaching in Cambodia. The collaboration involves SeeBeyondBorders, a non-government organization concerned with teacher development in Cambodia; the Department of Policy, a research and think-tank within the Ministry of Education, Youth and Sports; Phnom Penh Teacher Education College as provider of the new full 4-year degree course for primary teacher training and the Centre for Education and Lifelong Learning at National College of Ireland.

The research goals of the PIKT project may be characterized in terms of three broad areas of inquiry. First, to identify an appropriate framework to facilitate early grade Cambodian teachers to enhance their classroom practice. Second, to devise a workshop with associated resources to support teachers to enhance their classroom practice in line with the framework. And third, to further develop the practical research skills and capabilities of the collaborating partners so that the work can continue in the future. The 28-month project commenced in September 2022 and will continue through to the end of 2024.

The PIKT project is investigating how teachers in Cambodia and specifically the Battambang region, can be enabled to improve their teaching practice and engage in a path of professional development. It focuses on the tasks and challenges of building teacher professional identity; how teachers see themselves and how this may change over time. Teacher professional identity is a core enabler for a wider framework of teaching capability and competence. Effective teaching requires a disposition for enhancement of practice in a continuous cycle of professional improvement.

PIKT uses design-based research to develop and deliver customized workshops with associated materials, to enhance teacher professional identity and build teacher agency in implementing positive changes to classroom practice in primary schools.

As part of the design-based research process, 25 volunteer participant teachers were observed in class and this data, together with teacher interviews, and a ‘school in community’ profile, enabled the research team to design and deliver a culturally and contextually appropriate workshop for the teachers. We report on the project to date as it has completed the first iteration of design and workshop delivery.

Insights from PIKT contribute to our understanding of the essence of teacher professional identity and how, despite difficult circumstances, teachers in Cambodia are well motivated to improve the lives and education outcomes of their students. The research also provides new insights on what's happening in Cambodian schools as it fosters further collaboration between Irish, international, and Cambodian researchers.