These pages work best in web standards compliant browsers like Firefox and Safari.
Test your developmental intuition: Take the Sort Task Challenge
DiscoTests do what conventional tests can't do. They tell teachers how well students think about what they have learned, and help direct students to the material they are most likely to benefit from learning next. Click on one of the tabs below to learn more.
In 2002, eight physical science teachers in Springfield, MA asked a group of professors and staff from Hampshire College to help them do a better job teaching basic physics concepts like energy and waves, which their students found particularly difficult to learn. While some of us worked with teachers exploring new ways to teach these concepts, others set out to study how students learned them. The research team began by studying how children between the ages of 5 and 17 come to understand the physics of energy.
Based on this research and the contributions of other researchers and content experts, we were able to describe how students typically learn the energy concept (Dawson, 2006).
Equipped with this knowledge, we designed a new kind of assessment and scoring rubric for teachers. Rather than examining how well students define terms or apply formulas, this assessment focused on how well students think about energy, by asking them to describe what is happening to the energy of a ball in a variety of situations. This assessment not only allowed teachers to examine how their students understood the physics of energy, but to determine what each student was most likely to benefit from learning next as they constructed increasingly adequate conceptions.
Although the teachers liked the assessment and found the rubric intuitive, they expressed two concerns: (1) administering and scoring the assessment was too time-consuming, and (2) there weren't enough assessments of this kind.
The Disco initiative was launched to address these concerns.
The Disco mission is to bring high quality FREE online developmental assessments to 7th - 12th grade public school teachers all over the English-speaking world. (For information about using DiscoTests in research or in private school settings, contact us.)
All DiscoTests will be grounded in rigorous research into the way students learn particular concepts, and will focus on how well they use these concepts to reason about real-world issues and problems.
All DiscoTests will be designed so they can be accurately coded by teachers in 5 - 10 minutes, making it possible for them to assess student thinking at relatively frequent intervals throughout the school year.
Coding rubrics for all DiscoTests will be continually refined to reliably produce accurate assessments of student performance.
All DiscoTests will be accompanied by student performance reports that not only describe a student's current performance, but also what that student is most likely to benefit from learning next.
To the extent that it is possible, we will design DiscoTests that meet public school teachers' most urgent needs.
Each DiscoTest will be complemented with information about the pathways through which students generally learn the concept(s) addressed by that test, as well as links to information about learning activities that support understanding and reasoning.
No identifying student information will be collected on the Disco site, although we will collect limited demographic information.
The essays written by students will become part of a growing database of student performances that will be used by researchers to improve (1) our understanding of the way students learn, (2) curricula, and (3) assessments.
In the twentieth century, test developers made important advances in examining students' content knowledge. Today, most tests of content knowledge are fairer and more accurate than they were 100 years ago. But test developers have had less success when it comes to evaluating how well students think about what they know.
One concern of test developers (including classroom teachers) is accuracy. When students make choices on a multiple choice assessment, the person marking the test only has to determine if he or she has chosen the right letter. The student's response is either right or wrong. On the other hand, when students write a response, like an essay, that shows their thinking, someone has to make a judgment about the quality of the response. This raises many concerns about accuracy that do not arise in multiple choice assessments—one of the main reasons students' ability to reason about what they have learned is rarely evaluated formally.
A second concern is efficiency. Today's teachers have little time to mark individual papers. Multiple choice and short answer (right or wrong) items are quick and easy to score. However, in order for students to show their thinking, they must explain their answers. Scoring essay responses requires a much greater time investment than scoring multiple choice responses. Ask an English teacher.
Unfortunately, when we test students' content knowledge with conventional assessments, we don't learn as much as we need to know about what they understand. Research shows that as many as half of the students who get a multiple choice test item right cannot demonstrate understanding of the problem posed in the item (Schoenfeld, 2007). In this time of No Child Left Behind and high stakes, multiple choice testing, this is a serious issue. To the extent that we "teach to" these tests we under-prepare our students for the challenges of 21st century life by focusing too much on content and not enough on understanding.
DiscoTests increase the efficiency with which teachers can assess student understanding while increasing the accuracy of scoring.
We address efficiency in two ways. First, we put the entire testing process online. Once a teacher has registered his or her students on the Disco site, students simply sign in to take an assessment. (We do not collect identifying information about students.) Second, we optimize the scoring process. All assessments, whatever the topic, require students to write 5-7 short essay responses. The teacher "codes" each assessment, using a set of 3-7 pull-down menus per essay. The pull-down menus allow teachers to code individual assessments in 5 -10 minutes (once they are familiar with a particular coding scheme). Third, student performance reports are automatically generated, based on teachers' codes.
Scores on DiscoTests are more accurate than the holistic scores produced using conventional scoring rubrics, for two reasons. First, the coding system allows us to calculate a score based on 20-30 independent observations. Once coding is complete, a score is calculated with an algorithm that analyzes the pattern of codes across all of the essays. Scores that are based on multiple observations are more likely to be accurate than those based on a holistic assessment. This is why tests are generally made up of multiple items. Second, selecting options from the pull-down menus does not require a deep analysis of a student's performance. Rather, options address the specific way a given student writes about a particular concept in a particular essay. This specificity makes it more likely that different teachers would agree on a rating. In fact, we're finding that juniors and seniors in high school are able to code their own performances with a high degree of accuracy.
