Principal Investigator: Dr. Michael Russell & Joan Lucariello
Bridging the Gap: Applying Algebra Cognition Research to Develop and Validate Diagnostic Classroom Algebra Testlet
Currently used tests of mathematics achievement provide information about whether students solve a problem correctly, but they provide little or no information about how students solve problems, or about what kinds of misconceptions might have led to choosing the wrong answer. Consequently, these tests provide little guidance to teachers about how to focus their teaching to help students overcome their misunderstanding. The purpose of this project is to develop a set of computer-based algebra short tests-or testlets-to provide information about students' misconceptions about or understanding of specific algebraic concepts that will enable teachers to use that information to guide their instruction of those students.
In the first phase of this project, the researchers are developing a series of short testlets designed to identify students' abilities to solve particular types of algebra problems and whether they possess various specific algebraic misconceptions that have been identified by previous research as being fairly common. Phase 2 studies will examine whether testlets identify the misconceptions they are designed to identify. Multiple methods will be used in a series of studies involving eighth and ninth grade algebra students, including studies comparing testlet performance with predictions of teachers and with student classroom work, and pretest-posttest studies examining the effects of targeted feedback or instruction on testlet performance. Among these will be a series of two group studies (one for each targeted algebraic misconception) in which 20 students who are identified with the misconception based on testlet performance will be randomly assigned to one of two groups (Group A or B). Students in Group A will work with an individual tutor who will provide instruction over a short period of time; students in Group B will not receive any instruction during this time. After the instructional intervention for students in Group A is complete, all students Groups A and B will retake the testlet. Students in Group B will then work individually with a tutor and all students will be tested for a third time. Changes in students' testlet performance will then be compared between the two groups.
The final phase of the project will involve an examination of whether teachers can use testlets to identify misconceptions of their eighth and ninth grade students, and to respond appropriately. Researchers will instruct teachers in how to use the testlets, after which pretest-posttest gains in testlet performance will be compared across four groups of students: 1) those who do not demonstrate a specific misconception, 2) those who have the misconception and who receive no targeted instruction, 3) those who have the misconception and receive inappropriately-targeted instruction (i.e., targeted at the wrong misconception), and 4) those who have the misconception and receive appropriately-targeted instruction. The prediction is that students in the fourth group will show greater pretest to posttest improvement in testlet performance than will students in the other three groups.
The overall goal of this project is the development of a diagnostic tool (i.e., a set of computer-based algebra short tests) that can provide teachers with information that they can use to guide instruction for individual students and groups of students who share similar problems with algebraic concepts, thus making classroom assessment a more instructionally useful component of teaching and learning.
Florida State University
Principal Investigator: Dr. Carol Connor
Child Instruction Interactions in Early Reading: Examining Causal Effects of Individualized Instruction
Research suggests that early language and literacy skills have important influences on children's reading development. Children who begin school with weaker language and early reading skills are less likely to experience academic success over the course of their school career. Moreover, recent research indicates that the effects of instruction on reading depend on the specific early language and literacy skills that children bring to school. In other words, there are child-by-instruction interactions such that certain forms of instruction may be more effective for some children, while other forms are more effective for other children. If this is so, it strengthens the idea that the kind of reading instruction a young student receives should be tailored to the particular strengths and weaknesses of the individual student. The challenge to this idea is the difficulty that teachers face in trying to provide individualized instruction to an entire classroom of students. The purpose of this project is to create and evaluate an approach to professional development and technology use that is designed to enable teachers to provide effective individualized reading instruction that can enhance children early reading skills, and to examine the effects of individualized instruction on growth in students' letter/word recognition and passage comprehension skills.
The first year of this project is devoted to developing a training program to show teachers how to create and implement lesson plans based on individual students' developing skills, and to creating software and web-based materials for use by teachers that will analyze student assessment data and produces recommendations for individualized instruction. Following that, a study will be carried out in order to evaluate the efficacy of this professional development and technology support approach. Twenty first grade teachers from an economically diverse school district will be randomly assigned to receive the training and technology support for implementing individualized instruction, or to receive the district's typical professional development training. A variety of observational methods will be used to determine whether teachers who receive the professional development and technology support program are more effective in providing individualized instruction to their students than are teachers who participate in district-typical professional development. The effectiveness of the program in improving student literacy skills will be evaluated by comparing student performance on standardized measures of literacy skills across treatment and control classrooms. The overall goal of this project is to produce new professional development and technology support tools to help teachers provide the kinds of individualized reading instruction that may be useful in promoting all children's early reading achievement.
University of California, San Diego
Principal Investigator: Dr. Harold Pashler
Optimizing Resistance to Forgetting
The value of education depends not only on what is learned, but also on whether information, once learned, can be retained. Often the explanation for why students don't know something is not that they didn't learn it, but that they learned it and subsequently forgot it. The purpose of this project is to address the question of what instructional procedures foster long-term retention of information and skills, with a particular focus on factors that can be readily manipulated in almost every learning context, be it a classroom or a computer-aided instruction or distance-learning program.
The researchers in this project are adding a series of ten cognitive experiments to an ongoing research program examining factors that affect retention or forgetting of learned information. Two studies are designed to investigate the effects of different study schedules (i.e., different lengths of time between study sessions) on the retention of skills and knowledge, using vocabulary words taught to 8-10 year old children in one study, and people who visit an Internet site to learn about skin cancer detection in the other.
Two additional studies are designed to examine testing effects on fact and foreign language learning. Specifically, these studies will compare the effects of regular study activity (i.e., the simple presentation of information to be learned) versus using testing as a study method, and whether observed differences in learning vary depending upon the length of the retention interval.
One study will compare discovery learning with learning under more passive conditions. Specifically, this study will examine whether children and adults who discover knowledge for themselves are more likely to learn and retain that knowledge than are individuals who receive the same knowledge more passively. In this study, participants will be randomly assigned to one of five different learning conditions varying in the extent to which they require active hypothesis testing to discover an underlying principle, versus being given the principle directly, and learning will be assessed following a 30-day retention interval.
Two studies will focus on whether receiving intermittent versus consistent performance feedback affect learning and retention of information, and two additional studies will assess the differential effects on learning and retention of six feedback conditions varying in how long feedback is delayed and how long it is available once presented.
The tenth study will examine the effectiveness of an intervention based on findings from the earlier studies, focusing on material that is already being taught to students enrolled in early high-school grades—particularly content that is taught once and that that is not reviewed or practiced in other classes. Students will be randomly assigned to one of several review intervention conditions. The review session will take place at several possible delays after their classroom learning took place. Delay intervals and procedures for review will be developed based on earlier findings. After a delay of up to 18 months after the review session, a final test of retention will be given.
Finally, the researchers will use information obtained from these studies to develop a computer-based intervention designed to teach English vocabulary to learners such as high school students preparing for the SAT. Results from this work should provide new information relevant to classroom practice and instructional interventions designed to improve students' ability to retain learned information over extended periods of time.
University of Wisconsin — Madison
Principal Investigator: Dr. Brian Bottge
Advancing the Math Skills of Low-Achieving Adolescents in Technology-Rich Learning Environments
U.S. students consistently score lower than students in many other industrialized nations on international tests of mathematics, and many students score below basic levels on national assessments. Research suggests that improving the skills of students with learning disabilities (LD) or emotional disabilities (ED) is especially difficult. Studies showing that students with LD gain, on average, only one year of achievement in math for every two years they are in school indicate that considerable effort is needed to upgrade their basic math and problem solving skills. The purpose of this project is to implement, test, and refine the most effective way to teach both problem-solving and basic skills to low-achieving adolescents using a teaching method called Enhanced Anchored Instruction (EAI).
EAI uses a mix of video-based problems delivered on CD-ROM (called anchors) and hands on projects (e.g., building skateboard ramps, compost bins, or hovercrafts). Each anchored problem consists of several sub-problems embedded in a realistic and motivating context. Anchored problems require considerable time to solve, usually five to ten 60-minute class periods. Students must first define and understand the EAI problem, locate the relevant pieces of information for solving it, and then integrate this information into a solution that makes sense. In previous work, students have been found to identify with the video's main characters and to work hard on helping them solve their problem. The additional practice the applied projects afford students appears to help them understand the importance and benefits of learning math.
The current study builds on the researchers' past work by examining the relative effectiveness of EAI for improving students' mathematics achievement in two conditions. In the first condition, basic skills such as computation of whole numbers and fractions are explicitly taught before and during EAI as planned units of instruction. In the second condition, these same skills are taught informally as they are needed in student efforts to solve the mathematical problems presented during EAI. The researchers are carrying out this study with classrooms of students including both average achieving students and low achieving students, many of whom are likely to have LD or ED disability classifications. The participating middle and high schools serve diverse ethnic and socioeconomic student populations. In the study, thirty classrooms of students are being randomly assigned to receive EAI with either explicit or informal teaching of basic skills. The researchers are testing students' mathematical achievement before and after the instruction to measure how much they learn from the EAI instruction, and attitude surveys and classroom observation are being used to study the student learning process as students attempt to solve complex problems. Using this information, the research is refining EAI to make it a more effective instructional approach for low-achieving students. The results of this project should provide new information on how best to integrate instruction of higher and lower order math skills in ways that facilitate the mastery of basic skills while at the same time motivating students to learn mathematics by providing engaging problem-solving opportunities with real-world applications.