Multiple-choice exams measuring the breadth of students鈥� content knowledge have hardly disappeared.
But the creators of a leading international exam, the U.S. government agency behind 鈥渢he nation鈥檚 report card,鈥� and the makers of new classroom learning tools are all turning their attention to something new: digital simulations aiming to measure how students solve problems, communicate, and work with others.
Examples include simulations that ask students to collaborate with a computer avatar to keep a tankful of tropical fish alive, help a virtual village fix its broken water pump, develop a model of an atom, and help an online city reduce pollution while growing its economy.
鈥淚n the real world, there is often more than one route to the right answer, or more than one right answer,鈥� said Michael Barber, the chief education strategist for education publishing giant Pearson and a strong backer of the new assessment trend. 鈥淲ith simulations, you can have complex problems and allow different students to find different ways through.鈥�
The use of simulations for assessment isn鈥檛 new: For decades, for example, airline companies and the military have used simulators to train and screen prospective pilots. Many educators have also had a long-standing interest in more traditional forms of low-tech simulated performance assessments, in which students are required to complete a sophisticated task, such as designing and conducting a science experiment.
But questions of cost, reliability, validity, and scalability have been stumbling blocks in the past to widespread adoption of both simulations and performance assessments, said Eva L. Baker, a research professor in the graduate school of education and information studies at the University of California, Los Angeles. She said those same challenges remain with today鈥檚 emerging technologies.
鈥淲e all want to avoid the gee-whiz factor,鈥� said Ms. Baker, the director of the National Center for Research on Evaluation, Standards, & Student Testing, or CRESST. 鈥淲e have to make sure there is adequate evidence these things work and are not simply the next big thing.鈥�
Real-Time Feedback
Proponents maintain that digital simulations, which experts in the field say generally involve less narrative, design, and purely fun elements than digital games, can increase student engagement and provide real-time feedback to educators. They also believe the torrents of data that simulations generate can be mined for evidence of deeper learning鈥攏ot just whether a student selects the correct canned response, but whether he or she asks the right questions, uncovers all the necessary information to make a smart choice, and shares that information effectively with a partner.
鈥淜nowing the right answer is just half of what we should be looking for,鈥� said Peggy G. Carr, the associate commissioner of the assessments division at the National Center for 91制片厂视频 Statistics, which this spring is field-testing its new simulation-heavy technology, engineering, and literacy assessment with 20,000 U.S. 8th graders.
But while the TEL exam鈥攍ike the Program for International Student Assessment, or PISA, which will begin incorporating digital simulations in 2015鈥攊s a summative test that aims to provide an overview of student performance, the real power of simulations lies inside the classroom, said Christopher Dede, a professor of learning technologies in Harvard University鈥檚 Graduate School of 91制片厂视频.
鈥淚f we can develop better formative assessments that provide diagnostic information that is used to go back and change the course of instruction,鈥� Mr. Dede said, 鈥渨e can have a much greater impact.鈥�
Some groups are working on exactly that goal: Pearson, which is based in London and has U.S. headquarters in New York City, is playing a key support role in the development and rollout of SimCityEDU, a classroom adaptation of the classic urban-planning computer game. And PhET, a research and development laboratory at the University of Colorado at Boulder, is currently expanding the formative-assessment tools built into its 128 digital math and science simulations.
Barriers remain: It takes about six months and $50,000 to build a single PhET simulation, for example, and parents and privacy advocates are likely to worry about how all the student data that simulations generate is secured. On a practical level, the nation鈥檚 schools are likely to be occupied for the foreseeable future with the challenge of moving even more traditional assessments online.
鈥淚 don鈥檛 think this is going to change the face of education in a two- or three- or even a five-year period,鈥� said Mr. Barber of Pearson. 鈥淏ut I think in 10 years, it will be transformative.鈥�
Digital simulations are making their way into a wide variety of assessments, used for an equally wide variety of purposes. Four of the prominent players in this emerging field include:
1. National Center for 91制片厂视频 Statistics
How do you measure students鈥� ability to understand, use, and improve technology?
With 鈥渢ask-based scenarios鈥� online, believe officials of the NCES, which administers the National Assessment for 91制片厂视频al Progress.
In one such task on the agency鈥檚 new technology, engineering, and literacy exam, students are asked to assume the role of an engineer who must figure out why a village鈥檚 water well has stopped pumping water. The item gauges test-takers鈥� troubleshooting skills: Do they access and use the most relevant information? Can they quickly diagnose the problem?
鈥淚t鈥檚 intended to provide an assessment that鈥檚 more authentic for students,鈥� said Peggy G. Carr, the associate commissioner of the assessments division at the NCES, a branch of the U.S. Department of 91制片厂视频.
The exam is currently going through its second large field test, although no live administration has yet been scheduled. The NCES is investing in the new assessment, Ms. Carr said, in part because science, technology, engineering, and mathematics skills often cut across academic content areas and are untouched by subject-specific tests.
The reliability and validity of digital simulations remains a concern, but she said an earlier field test of the new exam helped: The NCES feels confident, for example, that it has identified the most efficient way for test-takers to troubleshoot the broken pump. Test-takers who pursue that path will get the highest score. 鈥淭here has to be a way to score these tasks such that you can put them on a scale with higher or lower levels of proficiency,鈥� Ms. Carr said.
2. Organization for Economic Cooperation and Development
A leading global assessment used to compare the reading, mathematics, and science skills of students in 70 countries will use digital simulations to measure students鈥� problem-solving ability鈥攂oth alone and with others鈥攂eginning in 2015.
In science, students taking the Program for International Student Assessment, or PISA, next year will encounter 鈥渆xperimental situations鈥� in which they must manipulate multiple variables in order to perform a real-world task, such as figuring out how to make the most efficient zeer pot (a clay refrigerator that doesn鈥檛 require electricity.)
PISA will also include in 2015 a new exam solely dedicated to measuring students鈥� collaborative problem-solving ability. In one sample item, the challenge is to keep a tankful of tropical fish alive. In order to succeed, students will need to interact with a computer avatar via a chat window to uncover necessary information, devise and implement a plan of action, and communicate the results of their experiment.
鈥淪imulations test students鈥� thinking skills and creativity in far greater ways than we can do with pencil and paper,鈥� said Michael Davidson, the head of the early-childhood education and schools division of the Paris-based OECD, which administers PISA. 鈥淚f countries agree that these are skills 15-year-olds should have, then having a summative assessment like PISA to determine if those skills are present is useful.鈥�
3. PhET Interactive Simulations Project
For some, digital simulations are less about determining students鈥� proficiency than about helping them learn.
鈥淥ur simulations are providing dynamic feedback so students can build an understanding of cause-and-effect relationships, develop mental models, ask questions, generate evidence, and engage in the scientific process at the same time they鈥檙e learning core content ideas,鈥� said Kathy Perkins, the director of PhET, based at the University of Colorado at Boulder.
Since 2004, the PhET team has created 128 interactive digital simulations in physics, biology, chemistry, earth science, and math. All told, Ms. Perkins said, the simulations have been used more than 45 million times, primarily in middle and high schools.
In one popular example, students can drag and drop protons, neutrons, and electrons into a simulated atom.
Using the PhET simulations for formative assessment requires a lot from teachers, who generally must infer for themselves how students are interacting with the tools, have the skills to get students talking about what they鈥檙e learning, and know how to adjust classroom instruction based on the resulting insights.
To ease that burden, PhET is working to expand its professional-development resources for teachers and working with external partners such as Pearson and the Princeton, N.J.-based 91制片厂视频al Testing Service to think about new tools and functions that can be added to the simulations.
鈥淲e want to create tools that the broader education community can use and leverage,鈥� Ms. Perkins said.
4. Pearson
Pearson has long been a titan in the standardized-testing industry, but officials from the London- and New York City-based company say they鈥檙e most excited about the potential of digital games and simulations to provide formative support inside classrooms. 鈥淚 think [simulations] have a place in summative assessments, but you have to take their limitations into account,鈥� said Kristen DiCerbo, a senior research scientist for the company.
Ms. DiCerbo has worked extensively with GlassLab鈥攁 team of researchers, designers, developers, and learning scientists making games for assessment, in which Pearson is a participant鈥攐n the development of the recently released SimCityEDU. A classroom adaptation of the classic urban-planning computer game, SimCityEDU鈥攚hich Ms. DiCerbo describes as more game than simulation due to the level of design, narrative, and just plain fun that is built in鈥攁ims to promote and assess students鈥� 鈥渟ystems thinking,鈥� or grasp of multiple overlapping cause-and-effect relationships.
As students try to reduce pollution while also improving the economy in a virtual city, their every digital action is tracked: 3,000 points of data or more in a 10-minute play cycle.
Ms. DiCerbo leads the effort to mine that information, searching for patterns that suggest evidence of learning. How to safely secure all that student data remains a sensitive issue, and there have been surprises鈥攚hat to make, for example, of the 5 percent to 10 percent of players who decide the most efficient way to meet SimCityEDU鈥檚 goals is simply to bulldoze big sections of the city.
