|Title:||Enhancing Learning and Transfer of Science Principles via Category Construction|
|Principal Investigator:||Kurtz, Kenneth||Awardee:||State University of New York (SUNY), Binghamton|
|Program:||Cognition and Student Learning [Program Details]|
|Award Period:||3 Years (7/1/2012-8/31/2015)||Award Amount:||$754,846|
|Type:||Development and Innovation||Award Number:||R305A120554|
Co-Principal Investigator: Andy Cavagnetto
Purpose: Poor academic performance has been linked to the tendency of students to focus on surface level characteristics of information in learning and when encountering new situations. To combat the often inaccurate and inaccessible concepts that result from this tendency, the researchers will develop an intervention based on an established psychological task known as category construction. The goal of category construction is to guide students in scientific thinking in much the same way that natural object categories (e.g., dog, chair) guide learners in making sense of everyday experiences. In this project, researchers will develop and test a set of tasks in which middle school students who are learning evolution concepts classify relevant science text passages written on note cards by organizing them according to common characteristics.
Project Activities: Using a series of six experiments and one pilot study designed to test the potential efficacy of the category construction intervention in seventh-grade classrooms, the researchers will examine whether students can learn about evolution concepts through the classification of relevant text passages written on note cards that can be organized according to common characteristics. Nearly 500 seventh-grade students will be involved in the project. The six experiments will be conducted in both laboratory and classroom settings and will be used to determine which features of the intervention support student learning and should be incorporated into the final intervention. A pilot study conducted in six seventh-grade classrooms during the final year of the project will assess the promise of the fully developed intervention.
Products: The products of this project will be a fully developed category construction intervention for seventh-grade students, evidence of the promise of the intervention, and peer-reviewed publications.
Setting: Experiments 1 and 2 will take place outside of the classroom in a middle school setting in New York. Experiments 3 through 6 and the pilot study will take place in life science classrooms in this same middle school.
Sample: Three seventh-grade teachers will provide feedback throughout the initial development of the materials and when integrating the materials into their classroom instruction. Between 450 and 500 seventh-grade students will participate across three years. Just over one-third of the students in the school district qualify for free and reduced lunch services. Currently 88 percent of the students are Caucasian, 6.3 percent are African American, 2.4 percent are Asian/Pacific Islander, 2.2 percent are Latino, and less than 1 percent are Native American.
Intervention: The intervention is based on the category construction task, which requires students to sort a set of six surface-dissimilar examples (short text passages describing evolution scenarios) into two categories or groups that share a common underlying principle. Each set will consist of three positive examples (demonstrating a target principle in the evolution unit of life science) and three negative examples (representing a coherent contrasting category or alternate principle). In order to complete the categorization task, participants must identify and apply the abstract principle that unites the three positive examples and learn how to link the generalization to its specific surface forms.
Research Design and Methods: The researchers will complete six experiments and one pilot study designed to test the potential efficacy of the category construction intervention in seventh-grade classrooms. In the initial year, the team will carry out two experiments. In Experiments 1 and 2, the researchers will establish, investigate, and evaluate a core training task based on category construction. During the second year, the researchers will use an iterative development process (where each experiment builds on the findings of the previous experiment) to integrate the intervention with ongoing instructional activity in the classroom. Using material from the teacher's current 4-week Evolution instructional unit, the studies will explicitly address the optimal form of this integration. Experiment 3 will evaluate the best instructional moment to carry out this task, at the beginning of instruction, at the end of instruction, or after a delay. Experiment 4 will evaluate whether incorporating the naming of the principle or generating an explanation of the principle improve student learning over and above the standard sort task. Experiment 5 will test the impact of feedback when students initially fail to correctly sort the examples. Experiment 6 will test the impact of repeated practice, and whether the practice is massed or spaced) on students' ability to learn from category construction. During the final year of the project, the researchers will conduct a pilot study to assess the potential efficacy of the fully developed intervention. Fidelity of implementation checks will be assessed using a fidelity protocol and audio recordings of teacher feedback.
Control Condition: Due to the nature of the study, each experiment has its own set of training conditions that serve as a comparison. The pilot study compares a natural baseline condition (which reflects what the teacher would have been doing instructionally with regard to the target principle and serving as a fair control in terms of time allocated and use of three sessions, as well as in terms of conforming to standard practices) to a prototype condition, which employs the optimized category construction task.
Key Measures: Dependent measures include both quantitative and qualitative assessments of concept understanding and transfer (that is, the ability to spontaneously apply the target principle to a test problem presented immediately or after delay, which lacks surface similarity to the training cases). These measures are collected in conjunction with the learning intervention, as well as integrated with classroom assessment (unit tests). Transfer performance will be measured in three ways: transfer tests at the end of each learning module (total instruction on a targeted principle), the teacher's actual unit test (Experiments 1–6), and an experimenter-designed delayed transfer test one week after the evolution unit. These tests will address basic mastery of the eight target principles and transfer questions that require application of the principles. Test items will either be pulled directly from or adapted from the Conceptual Inventory of Natural Selection and researcher-developed open-response items.
Data Analytic Strategy: Data analysis consists of a scoring procedure followed by significance testing using analysis of variance, t-tests, and chi square tests. Analysis will examine whether consistent classroom use of category construction in the evolution unit with the additional supports of feedback, naming, and practice will produce better understanding and transfer of evolution concepts relative to the comparison conditions.
Kurtz, K. J. Cavagnetto, A. R., Honke, G., Conaway, N. B., Patterson, J. D., Marr, J. C. and Tao, Y. (2014). Optimizing the Category Construction Task to Promote Learning and Transfer of Knowledge in Classroom Instruction. In 36th Annual Conference of the Cognitive Science Society. Quebec City, Canada: Cognitive Science Society.