Performance Assessments: More Building, Less Bubbling

Since my last post, I've been stewing a little on the value that we can add to school districts in this current educational climate. Over the past 10 years or so, there has (rightly) begun to be an intense focus on gathering data on what students really know and where their gaps in comprehension may lie. Our current system of formative and summative assessment is heavily based on periodic benchmark tests and annual, http://www.blogger.com/img/blank.gifhigh-stakes exams. In the end, it sadly becomes about students filling in the correct bubbles and using test taking strategies like elimination to help them figure out the answers that they know must lie somewhere between the letters A and E on the page. Useful, because they will have to take standardized tests for the rest of their lives. But not so useful when it comes thttp://www.blogger.com/img/blank.gifo assessing how well they might perform when they are asked to build and test a robot or a simple radio in an engineering class. Especially thinking about kids in our underserved neighborhoods, it seems crazy to expect them to make the transition from bubbling to building in a few months between high school and college. Enter SimInsights and the creation of computer based performance assessments.

Here's the lowdown on performance assessment. Basically it's a far superior form of understanding where a learner is in their development, but until now, it's taken a lot of work on the creation and the actual evaluation side to make it something that can be done on a large scale. And speaking of scale, the SCALE group at Stanford School of Education has done some great work in documenting pilot studies of performance assessment. Instead of bubbling in answers, we can actually have students placed in a virtual environment where they have to demonstrate their proficiency with a topic by building a model or testing a hypothesis by creating an experiment and executing it. This type of system could really add value to a student's learning, and at the same time, provide bucketloads of useful data about what that student can actually DO with the information they have learned. Because we want to move towards a system with more building than bubbling.