Delivering the keynote address at the 2012 Strata conference, Steve Schoettler emphasized how learning analytics can be used to improve education -- a process that has been slow to take effect given the lack of data application.
Schoettler began his talk by pointing out that educational data collected by the federal government takes two years to process. For instance, a chart depicting U.S. average reading scores released in 2011 shows data only up to 2009. By not making use of data, there is no system in place for generating continuous feedback -- a detail that has stalled educational improvement.
“If we’re going to bring data to education, we have to know what to measure, and the most important thing to measure is the student, the learner -- that’s the center of education processes,” Schoettler says.
Today, students are evaluated through standardizes tests that measure skills and understanding. It's been shown, however, that test scores are often misused in policy decisions. To add to that, student scores are not representative of the whole picture, and fail to take into account complexities such as cognitive ability, working memory, multiple intelligences, learning style, family background, etc. -- all of which are necessary to measure in order to improve education, according to Schoettler.
Models that include these qualities this are used in social gaming to understand motivation, and in social networks to recommend books or people you may know. This kind of personalization could be applied to education, Shoettler says.
One major study Schoettler cites concludes that the most important factor affecting student achievement is feedback, which can be generated using formative assessments and then tailoring instruction to each student’s needs based on the results. These methods have already been used by schools that have doubled school performance.
Still, such ideas are time-consuming, and not every teacher has the tools or time necessary to implement these strategies, Schoettler admitted.
Ideally, he says, technology could be used to spread these ideas to every school in the country and make the relevant techniques available nationwide. Also, big data could pave the way for personalized learning programs by employing the same algorithms that are already used to recommend books.