More Like this

1. Comparing the Use of Two Different Approaches to Assess Teachers’ Knowledge of Models and Modeling in Science Teaching

   https://doi.org/10.3390/educsci13040405  

   Carroll, Grace; Park, Soonhye (April 2023, Education Sciences)

   Science teacher knowledge for effective teaching consists of multiple knowledge bases, one of which includes science content knowledge and pedagogical knowledge. With the inclusion of science and engineering practices into the national science education standards in the US, teachers’ content knowledge goes beyond subject matter knowledge and into the realm of how scientists use practices for scientific inquiry. This study compares two approaches to constructing and validating two different versions of a survey that aims to measure the construct of teachers’ knowledge of models and modeling in science teaching. In the first version, a 24-item Likert scale survey containing content and pedagogical knowledge items was found to lack the ability to distinguish different knowledge levels for respondents, and validation through factor analysis indicated content and pedagogical knowledge items could not be separated. Findings from the validation results of the first survey influenced revisions to the second version of the survey, a 25-item multiple-choice instrument. The second survey employed a competence model framework for models and modeling for item specifications, and results from exploratory factor analysis revealed this approach to assessing the construct to be more appropriate. Recommendations for teacher assessment of science practices using competence models and points to consider in survey design, including norm-referenced or criterion-referenced tests, are discussed. 

   more » « less
2. Long-term data reveal greater intertidal oyster biomass in predicted suitable habitat

   https://doi.org/10.3354/meps13949  

   Smith, RS; Hogan, S; Tedford, KN; Lusk, B; Reidenbach, MA; Castorani, MCN (February 2022, Marine Ecology Progress Series)

   Habitat suitability models have been used for decades to develop spatially explicit predictions of landscape capacity to support populations of target species. As high-resolution remote sensing data are increasingly included in habitat suitability models that inform spatial conservation and restoration decisions, it is essential to validate model predictions with independent, quantitative data collected over sustained time frames. Here, we used data collected from 12 reefs over a 14 yr sampling period to validate a recently developed physical habitat suitability model for intertidal oyster reefs in coastal Virginia, USA. The model used intertidal elevation, water residence time, and fetch to predict the likelihood of suitable conditions for eastern oysters Crassostrea virginica across a coastal landscape, and remotely sensed elevation was the most restrictive parameter in the model. Model validation revealed that adult oyster biomass was on average 1.5 times greater on oyster reefs located in predicted ‘suitable’ habitat relative to reefs located in predicted ‘less suitable’ habitat over the 14 yr sampling period. By validating this model with long-term population data, we highlight the importance of elevation as a driver of sustained intertidal oyster success. These findings extend the validation of habitat suitability models by quantitatively supporting the inclusion of remotely sensed data in habitat suitability models for intertidal species. Our results suggest that future oyster restoration and aquaculture projects could enhance oyster biomass by using habitat suitability models to select optimal site locations. 

   more » « less
3. Measuring Agenda Setting in Interactive Political Communication

   https://doi.org/10.1111/ajps.12653  

   Rossiter, Erin L. (September 2021, American Journal of Political Science)

   null (Ed.)

   Although strategies exist to measure actors' efforts to set policy, media, and lawmaking agendas, political scientists lack a method for identifying and accurately measuring another form of agenda setting that lies under the surface anytime two people talk. Within interactions, such as debates, deliberations, and discussions, actors can set the agenda by shifting others' attention to their preferred topics. In this article, I use a topic model that locates where topic shifts occur within an interaction in order to measure the relative agenda-setting power of actors. Validation exercises show that the model accurately identifies topic shifts and infers coherent topics. Three empirical applications also validate the agenda-setting measure within different political settings: U.S. presidential debates, in-person deliberations, and online discussions. These applications show that successfully setting the agenda can shape an interaction's outcomes, demonstrating the importance of continued research on this form of agenda setting. 

   more » « less
4. Auto-Encoding Variational Bayes for Inferring Topics and Visualization

   Pham, Dang; Le, Tuan (January 2020, Proceedings of the 28th International Conference on Computational Linguistics)

   Visualization and topic modeling are widely used approaches for text analysis. Traditional visualization methods find low-dimensional representations of documents in the visualization space (typically 2D or 3D) that can be displayed using a scatterplot. In contrast, topic modeling aims to discover topics from text, but for visualization, one needs to perform a post-hoc embedding using dimensionality reduction methods. Recent approaches propose using a generative model to jointly find topics and visualization, allowing the semantics to be infused in the visualization space for a meaningful interpretation. A major challenge that prevents these methods from being used practically is the scalability of their inference algorithms. We present, to the best of our knowledge, the first fast Auto-Encoding Variational Bayes based inference method for jointly inferring topics and visualization. Since our method is black box, it can handle model changes efficiently with little mathematical rederivation effort. We demonstrate the efficiency and effectiveness of our method on real-world large datasets and compare it with existing baselines.", 

   more » « less
5. Auto-Encoding Variational Bayes for Inferring Topics and Visualization

   Dang, Pham; Le, Tuan (December 2020, Proceedings of the 28th International Conference on Computational Linguistics)

   null (Ed.)

   Visualization and topic modeling are widely used approaches for text analysis. Traditional visualization methods find low-dimensional representations of documents in the visualization space (typically 2D or 3D) that can be displayed using a scatterplot. In contrast, topic modeling aims to discover topics from text, but for visualization, one needs to perform a post-hoc embedding using dimensionality reduction methods. Recent approaches propose using a generative model to jointly find topics and visualization, allowing the semantics to be infused in the visualization space for a meaningful interpretation. A major challenge that prevents these methods from being used practically is the scalability of their inference algorithms. We present, to the best of our knowledge, the first fast Auto-Encoding Variational Bayes based inference method for jointly inferring topics and visualization. Since our method is black box, it can handle model changes efficiently with little mathematical rederivation effort. We demonstrate the efficiency and effectiveness of our method on real-world large datasets and compare it with existing baselines. 

   more » « less