Cognition and Student Learning Education Research

Current Section
 Institute of Education Sciences Home
FY 2003 Awards
Archived Information

Carnegie Mellon University
Principal Investigator: John Anderson
The neural markers of effective learning

Many students currently struggle with algebra, yet mastering algebra is considered crucial to pursuing a career in mathematics or science. Computer-based tutoring programs have had some success in helping students learn algebraic procedures but have been less successful in teaching students how to solve algebra word problems. The purpose of this project is to improve a computer-based algebra tutor, using both behavioral and brain imaging techniques.

In the first phase of the project, the researchers are designing a 5-hour algebra unit focused on conceptually challenging word problems and adding it to a successful 9th grade algebra cognitive tutor program. Following that, the researchers are examining various markers of successful learning in college students, using brain-imaging techniques (i.e., functional Magnetic Resonance Instrument (fMRI)) to observe the learners' brain activity when insight and deep understanding are achieved. The researchers are then using brain imaging techniques (fMRI) to observe the brain activity of high school students as they solve problems with the algebra computer tutor. The results from this study will be used to further improve the computer tutor. Finally, the researchers are testing the efficacy of the revised computer tutor with classrooms of racially and socioeconomically diverse 9th grade students. The researchers are comparing the effects on learning with classrooms of students randomly assigned to use the original algebra tutor, the modified algebra tutor, or classroom discussion and practice of word problems. At the conclusion of this research project, this team will have a modified tutor that helps students gain the skills necessary to be flexible word problem solvers in the context of algebra.

Carnegie Mellon University
Principal Investigator: David Klahr
From cognitive models of reasoning to lesson plans for inquiry

International comparisons of student achievement in science show that while U.S. students at fourth grade are among the best in the world, they are only average at eighth grade, and even lower at twelfth grade. Some have attributed this decline to the incoherent, unfocused, and undemanding quality of existing middle school science curricula, and to the lack of emphasis on reasoning and problem solving in science teaching. The purpose of this project is to develop, implement, and assess cognitively-grounded lesson planning methods to improve science achievement during the middle grades.

The researchers are carrying out a three phase design experiment in 5th through 8th grade classrooms in four urban schools serving a predominantly African American population of students, most of whom come from low income families. In the first phase, the researchers are studying how science education is currently being delivered by examining existing state science standards, science curricula, teaching practices, and high-stakes tests and seeing how they relate to one another. The researchers are using content and task analyses of educational materials, ethnographic field observations of science classes, and videotaped lesson analyses. In the second phase, they are carrying out a researcher-teacher design study, in which they develop lesson-planning methods and use those methods to guide their teaching of science in regular classrooms, evaluating the potential efficacy and feasibility of the approach. In the third phase, they are using a professional development process to help regular teachers learn to apply the new lesson planning methods, with the goal of implementing sustainable changes in the teachers' teaching practices and impacting students' science achievement. The researchers will measures changes in participating teachers' instructional practices and their students' academic achievement in science.

Columbia University
Principal Investigator: Janet Metcalfe
Study enhancement based on principles of cognitive science

Research has shown that success in learning can be improved by controlling various factors in learning situations, such as differences in the spacing of learning opportunities and having answers generated by the students instead of having the answers presented to them. The purpose of this project is to examine how certain components of cognitive activity influence children's study behavior and related academic performance, focusing primarily on the effects on student learning of errors made in the course of learning activities.

The researchers are carrying out a series of eight studies focusing on the learning of children involved in an educational study enhancement program designed to improve the academic performance of diverse inner-city sixth and seventh grade children at high risk for academic failure. The intervention is a flexible computer-based study program based on principles of cognitive science and designed to target and improve memory and learning of vocabulary items. In the first two studies, the researchers are varying the students' learning experiences to investigate whether student learning is adversely affected when students make mistakes, and whether those effects vary with the learners' degree of confidence in their own answers. The third study is designed to compare the effects on learning of observing someone else making errors compared to making errors oneself. The fourth study involves the role of the timing of feedback and errors on learning. The fifth and sixth studies focus on learning in instances in which students give erroneous answers that they are highly confident are correct, and on the students' level of attention in these instances. In the seventh study the researchers are looking at the effects of different spacing patterns on learning, and in the eighth study they are investigating how excluding items the students have already learned influences their long-term knowledge of those items.

In short, these studies are determining whether making errors during learning helps or hinders mastery of the words to be learned and identifying the most effective use of study time in order to ensure that students master an entire set of vocabulary words. The investigators will incorporate the results of these experiments into a computer-assisted program that will support the learning of science, social science, and advanced English vocabulary terms of at-risk middle school children.

George Mason University
Principal Investigator: Robert Pasnak
Increasing learning by promoting early abstract thought

The ability to learn depends in part on the ability to apply abstract principles in a wide variety of contexts. Two particular abstract principles - learning to figure out which object in a group is unlike the others (the oddity principle), and knowing how to insert an appropriate object into a pre-given series of objects (seriation) - are forms of abstract thinking that are especially important in kindergarten as building blocks to greater school success. The investigators theorize that knowledge of these two principles promote abstract thinking and should enhance the acquisition of academic content knowledge (e.g., reading and math). The purpose of this project is to measure the impact of using an education intervention designed to enhance these two particular forms of abstract thinking in young children on those children's learning and achievement.

The research team is using an experimental design involving 24 kindergarten classrooms of students in an urban school system with an ethnically diverse population of children from low-income families. In each classroom, the 16 five-year-olds who score the lowest on the Primary Arithmetic and Language Scale (PALS) in their classrooms are randomly assigned to one of four groups. Four students are assigned to the experimental group, who receive explicit small group instruction and practice for 15 minutes per day on the oddity principle and inserting objects into series. Group Two receives mathematics instruction during the allotted time period; Group Three receives reading instruction; and Group Four receives instruction in the arts. The researchers are comparing the learning outcomes for the students in the four conditions using several different measures, so that they can compare student learning of the abstract thinking principles, and of their learning in mathematics and reading, to see which forms of instruction are more effective for attaining which learning outcomes.

University of California, Los Angeles
Principal Investigator: Keith Holyoak
A multidisciplinary study of analogical transfer in children's mathematical learning

Mathematics education in the United States does not enable a substantial number of students to meet state achievement standards for proficiency in mathematics, and many school systems lack effective programs for teaching students to think conceptually about novel or difficult problems. Research suggests that analogical reasoning plays an important yet little understood role in teaching mathematical thinking and learning. The purpose of this project is to explore ways to improve children's learning of mathematics through teachers' use of analogy in instruction. The researchers in this project are using both previous studies of typical U.S. classroom activities and laboratory experiments to study the instructional use of analogy as a mechanism for enhancing children's learning.

As a first step, the research team is analyzing videotapes of 25 mathematics lessons from a randomly selected sample of eighth-grade American classrooms to examine how mathematics teachers currently use analogies in their teaching, and how students rely on analogies in their learning. Based on those analyses, the researchers are then carrying out four experiments in which middle school students are randomly assigned to view videotaped mathematics lessons reflecting various aspects of the use of analogy in instruction. In the first study, students will be tested before and after they watch videotapes in which the teacher either uses an analogy explicitly while teaching a lesson about dividing fractions, or leaves the analogy unmentioned. The second experiment is designed to investigate the effects of students watching a student who is involved in developing a relevant analogy for solving a problem, and the third experiment is the same except for the addition of interviews designed to investigate what students have learned about the analogy involved. The fourth experiment is designed to examine student learning of different kinds of analogies. In all four experiments students are tested immediately after watching the videotape and after a period of time, to see what they have learned and how well they have retained it. If the use of analogies improves student learning, the researchers will develop workshops for teachers in the participating schools on the use of analogies in teaching mathematics.

University of Illinois at Chicago
Principal Investigator: Keith Theide
Improving monitoring accuracy improves learning from text

Theoretical models of self-regulated learning suggest that people monitor their learning as they study, using the information from their monitoring to guide their studying (e.g., decisions about what material to restudy or when to stop studying), which in turn affects learning. According to these models, accurate monitoring by readers is critical to the learning process. Research indicates that accuracy increases when individuals reread texts, write summaries, or generate a list of keywords for texts before rating their comprehension of texts. Improved accuracy leads to better regulation of study, which ultimately leads to improved reading comprehension. The purposes of this project are to study factors that influence the accuracy of readers' monitoring of their own comprehension, and to develop interventions to strengthen comprehension by improving that accuracy.

The research team is carrying out five studies using a variety of research methods to investigate the self-monitoring process in a diverse sample of university students, including a focus on the comprehension monitoring of struggling readers. The first study is designed to examine if summarizing passages immediately after reading them, summarizing after a delay, or not summarizing at all affects reading comprehension. The second study is testing whether the timing of summarizing what one has read affects reading comprehension. In the third, fourth, and fifth studies, the investigators are comparing the effects of different monitoring strategies on reading comprehension. With the results of these studies, the investigators will be able to identify key strategies for improving reading comprehension that can be incorporated into a reading comprehension intervention.

University of Maryland, College Park
Principal Investigator: Thomas Nelson
Computer-assisted instruction for learning and long-term retention based on recent cognitive and metacognitive findings

Although public interest in the use of computer-assisted instruction is strong, available computer tutoring programs generally do not make use of the most recent developments in technology and research on human learning, nor are they always tested in typical educational settings. The focus of this project is to develop a new model computer tutor to support learning and long-term retention of second-language vocabulary for students of various ages and grade levels.

The researchers are seeking to improve an existing computer tutor model for second language vocabulary in several ways. The new model builds upon prior research about various factors that influence learning and is designed to individualize instruction and testing based on the user's successes and failures on previous test items. Factors taken into account by the new design include the role of short term memory, the user's perception of the difficulty of the items, the difficulty of the items for a previously tested sample of learners, and the amount of time the user takes to recall the items. The research sample for the initial development of the model includes university students of all races and family income levels. The researchers are then conducting studies designed to evaluate the efficacy of the new computer model in a wide range of educational settings, which will include students in primary schools, middle schools, and university foreign-language courses. The researchers are randomly assigning students at each of these levels to use either the new computer model or else the baseline model from which the new model was developed, to see if the use of the new model produces significantly better student learning.

University of Pittsburgh
Principal Investigator: Erik Reichle
Lapses in meta-cognition during reading: Understanding comprehension failure

Virtually everyone has occasionally had the experience in reading of "zoning out", that is, of having their eyes continue to move across the page while their mind is elsewhere. Little research has directly examined the role that such "zoning out" experiences play in reading comprehension, and whether whatever negative consequences this experience leads to might be addressed by some sort of instructional guidance. The purpose of this project is to refine and validate a recently developed theoretical perspective for investigating zoning out during reading, and study the pedagogical implications of these occurrences.

The researchers are carrying out a three phase project with a series of studies in each phase, with undergraduate students as the subjects of the studies. In the first phase the researchers are studying how to induce and measure zoning out in laboratory settings. They are examining what text materials increase the frequency of zoning out; whether zoning out leads to comprehension failure; and whether indirect measures, including changes in eye movements during reading (as measured by an eye-tracking device) and alterations in brain activity (as measured by an electroencephalograph (EEG)) can be used to measure zoning out as well as self-reporting or questioning by an observer. In the second phase of the project, the researchers are carrying out studies of various cognitive factors associated with zoning out, including the role of the subject's lack of self-awareness while zoning out, their proneness to distraction, and their working memory capacity. In the third phase of the project, the researchers are exploring the pedagogical implications of their findings by studying whether or not mindful reading techniques or practice in the use of attention reduce zoning out, and whether the frequency of students' zoning out serves to predict poorer performance in college.

University of Wisconsin, Madison
Principal Investigator: Arthur Glenberg
Training indexing to enhance meaning extraction in young reader

Thirty-eight percent of U.S. 4th graders cannot read and understand a paragraph in a children's book, and the percentages for struggling minority students and students from low-income families are even higher. In this project the researchers are using advances in cognitive science and the study of language development to create and test an intervention to enhance young children's reading comprehension. The intervention is based on the Indexical Hypothesis, which theorizes that good readers associate words and phrases with objects and actions in the environment or mental images of objects and actions and use their ideas of these objects and images to make sense of what they are reading. The intervention is designed to teach children how to perform indexing, first with objects, and then with mental images. Children read texts describing a toy scenario arrayed in front of them (e.g., a farm scenario with a barn, animals, tractor, etc.), and, after reading a sentence, they manipulate the toys to correspond to the sentence. Following this manipulation training, children then learn to imagine manipulating (IM), in which they picture in their minds the objects and actions that a given sentence refers to. Training on the Indexical Hypothesis makes explicit that part of the reading process that requires the reader to link written words to what the words mean.

The researchers are carrying out four controlled experiments with random assignment to test the Indexical Hypothesis and demonstrate the intervention's efficacy with a culturally diverse sample of first and second grade students. In the first experiment the researchers are comparing students' ability to perform an IM task depending on whether their preparation for the task includes physical manipulation of objects, seeing those objects without touching them, or having those objects described to them verbally. The second experiment is designed to determine whether IM training increases students' ability to use their reading skills to carry out a practical task, such as following written instructions to build a particular object out of blocks. In the third experiment, the researchers are working with experimental classroom teachers to integrate IM into their curricula for half a year, and then comparing these students' reading comprehension skills with those of students in classrooms who continue to receive regular instruction. The fourth experiment involves following the progress of the children who participated in the third experiment for another half year to assess whether the IM intervention has enduring effects.

Washington University
Principal Investigator: Henry Roediger
Test-enhanced learning

The goal of this project is to examine a new approach to learning, memory, and comprehension of text material. The Test Enhanced Learning (TEL) approach views tests as opportunities for learning, instead of holding the traditional view of tests as assessment techniques. Prior laboratory research shows the power of retrieval of material during tests in enhancing later retention. In this project the researchers are moving toward a systematic application of this approach to education practice, investigating the power of tests to enhance learning of text material.

The researchers are carrying out five experiments with university students as subjects and using materials that come from actual science courses. The researchers are conducting three laboratory experiments to examine the power of tests to affect later retention on either the same type of test or a different type of test, relative to appropriate controls. They are also analyzing whether multiple tests produce larger benefits than single tests, and whether the spacing of intermediate tests affects performance on a final test. In two experiments the researchers are manipulating the testing in a regular university course, with different types of tests (e.g., multiple choice, short answer) and different schedules of testing being manipulated within-subjects during the course. Thus, in the final experiments they are applying their research to actual educational practice within college courses. The researchers are aiming to identify ways in which testing can be used to improve student learning.

Yale University
Principal Investigator: Robert Sternberg
Understanding students' mathematical competencies: An exploration of the impact of contextualizing mathematical problems

Technology-driven societies require a high level of mathematics and science proficiency in the general population, and so deficiencies in mathematical competence can seriously limit students' educational opportunities and prospects for future employment in such societies. The purpose of this project is to examine whether making mathematical problems appear more practical (that is, related to the context of everyday life), will improve students' performance on those problems and their mathematical knowledge of how to solve them, compared to problems that appear more academic (that is, abstracted from any everyday context). The researchers will investigate which of the factors that contribute to making mathematical problems seem more practical also contribute to better student performance on those problems, as well as how to shape instruction to take advantage of those factors.

The researchers are carrying out two series of studies with fourth and sixth grade public school students in an area serving a large proportion of low-income minority families. The first series of nine studies involve isolating various features of what makes a problem appear 'practical' to students tested on those problems, such as having physical objects to manipulate or being rewarded for the right answer, to determine whether those factors are related to better student performance. Using the results of these studies, the researchers will create a pool of test problems that reflect the presence or absence of those factors that contribute to giving problems a significantly practical context. In the second series of studies, classrooms of students will be randomly assigned to teachers using standard instructional materials or teachers using instructional materials designed to present mathematical problems in practical contexts, and the previously designed test problems will be used to determine the outcome. The topics covered by the instructional materials in these two studies will be equivalent fractions, which is relatively difficult to present in practical contexts, and measurement, which is relatively easy to contextualize.

Print this page Printable view Send this page Share this page
Last Modified: 05/04/2005