Paper Summary
Share...

Direct link:

Improving Middle School Science Instruction With Analogical Reasoning

Sat, April 18, 2:45 to 4:15pm, Sheraton, Floor: Second Level, Huron

Abstract

Purpose/Theoretical Framework

In a large, Institute of Education Sciences (IES)-funded, randomized-controlled-trial study, we integrated cognitive principles of learning into two different middle-school science curricula across four different school districts. The cognitive science-based intervention (CogSci) incorporated three major components (case comparisons Gentner et al., 2003], visualization exercises [Cromley et al., 2013] and spaced testing in the form of daily warm-up questions and repeated/delayed questioning on quizzes and tests [Roediger & Karpicke, 2006]) that were interleaved into each unit.
As an example, before instruction explaining the characteristics of all members of the plantae kingdom, students compared individual members of the kingdom to recognize their commonalities and to distinguish commonalities from the characteristics unique to specific members of the kingdom. When presented with a new organism within the plantae kingdom, students then could have transferred their knowledge of those commonalities and have had appropriate expectations as to the characteristics of this new organism.

Method

Four urban school districts in the northeastern and southwestern US granted us access to 6th-8th grade teachers’ time (n = 477) for professional development, classroom visits (n = 298 schools; see Table 1.1), and integration of our materials into their usual curriculum. The largest district’s existing curriculum was Holt’s text-based units centered around the biological and earth sciences (Cells and Introduction to Matter). The other districts were using the Full Option Science System (FOSS) centered around hands-on activities involving comparable subject matter (Life History and Weather & Water).
We modified existing curricula by integrating our CogSci tasks, but the original curriculum remained in place. Teachers received binders that included our modified materials (e.g., Powerpoints, worksheets) and explained their integration, calendars of when to use each, descriptions of activities, intended outcomes, and section post-tests. Teachers were randomly assigned within districts to one of three conditions: CogSci professional development, Science content-only professional development, or a business-as-usual control group. Unit post-tests were mailed to teachers toward the end of the four- to nine-week term (depending on the unit). Student (n = 43,076) and teacher demographics were also collected from districts, states, and teacher surveys. HLM models of students’ performance were fit to determine whether participation in the cognitive intervention predicts scores on post-unit tests.

Results

Results are shown in Table 1.2 and Figure 1.1: students in classrooms with >90% under-represented minorities gained little from the cognitive science intervention, whereas students (both White/Asian and URMs) in classrooms with <90% under-represented minorities gained significantly from the CogSci intervention. These analyses controlled for prior achievement, so the significant differences between higher-URM and lower-URM classrooms must be accounted for by other context effects such as time on task, teacher expectations, etc. Neither group benefited from the Content-only PD condition, compared to the control condition.

Significance

Results show promise for a large-scale application of cognitive-science based interventions within classrooms and provide helpful insights into how to organize subsequent interventions. We also highlight concerns when implementing and assessing cognitive-based interventions within classrooms.

Authors