Current information about this program can be found under the Education Research program.
Principal Investigators: William Whitten & Mitchell Rabinowitz
Guided Cognition of Unsupervised Learning
Purpose: Teachers generally assign homework to help students learn new information and procedures. From the perspective of cognitive scientists, the value of homework assignments for improving student learning depends on how the assignments are structured. The purpose of this project is to test strategies for structuring homework assignments in ways that will increase the likelihood that homework will involve productive study of new material. The researchers intend to create theory-based guidelines to help teachers and instructional designers construct more effective homework assignments. The overall goal is to determine the conditions under which students adopt cognitive strategies—such as summarizing, asking questions, and considering different answers to questions—when engaged in unsupervised study.
Setting: The schools are located in a suburban area located in the Mid-Atlantic United States.
Population: Participants include middle and high school English and science students from low-to-average and Advanced Placement classes. The students are white, Asian, Indian, Middle-Eastern, and African American. Some students are recent immigrants and are classified as English language learners. Socioeconomic backgrounds range from lower middle to upper class.
Intervention: A new method to improve the effectiveness of unsupervised learning (e.g., homework) is introduced and evaluated in this project. This method, called guided cognition, structures study tasks to guide the learner to engage in specific, observable cognitive events (e.g., drawing a diagram, listing multiple approaches to solving a problem, listing specific evidence that support conclusions). The idea is to make the thinking process more visible to the student. These cognitive events are hypothesized to elicit underlying cognitive processes that have been shown to facilitate learning in laboratory-based experiments. Unlike traditional homework, students completing guided cognition homework are provided with questions that are designed to elicit specific cognitive events. The researchers have developed these questions based on observations of the types of strategies teachers use to elicit thoughtful responses from their students. For example, in the context of classroom instruction, teachers often ask students to summarize a passage or illustrate visually a principle described in a text. The research team has identified a set of strategies that teachers use and adapted them to create questions and instructions that can be incorporated into homework instructions. Whereas a traditional homework question might be “Why does Macbeth visit the witches?,” a guided cognition question might be “Give two opposite but potentially valid reasons for this visit by Macbeth to the witches.” The researchers hypothesize that incorporating guided cognition questions into homework will improve student learning.
Research Design and Methods: The effect of guided cognition for students with different ability levels is being assessed through a series of 10 experiments in which students are randomly assigned to experimental or control conditions. Experiments are addressing time-on-task, learning-payoff-per-unit-time, novelty effects, guided cognition study without prior teaching, the relative effectiveness of specific cognitive events for facilitating learning, the extension of the paradigm to different subject matter and to different age groups, and transfer of learning as measured by students’ abilities to construct learning strategies that include previously experienced cognitive events.
Control Condition: In the control condition, students study in a controlled but unsupervised setting and are provided with traditional verbal study questions.
Key Measures: Learning is being assessed by unannounced quizzes on the content studied. Quizzes will not impact student grades. Student opinion data is being gathered via survey.
Data Analytic Strategy: Simple analysis of variance techniques are being used to examine student performance on the quizzes as a function of participation in the experimental or control condition.