Khan Academy and Altruism

Some years ago, while I was working for an educational software producer, a teacher/customer suggested a change that would improve the user interface and educational potential of our product. I agreed with the customer, and indicated that I had been advocating for that same change. A company Vice President who overheard the conversation later said, "Our product works well enough as is. It would cost too much to make that change and we don't have room in our product development cycle to do it." Another VP commented, "We're about profit, not altruism." This event was a sure indicator to me that I was in the wrong line of work: I was too much of a teacher not to be concerned about improving the educational value of our product. Within a year, I was back in the classroom.

However, occasionally, you will find something inspiring and truly altruistic. The Khan Academy (khanacademy.org) is a splendid example. Salman Khan has produced over 2,000 tutorials on topics such as math, science, economics and civics that he provides for free on his website. He started small by producing tutorials for some cousins who were struggling a bit in school. Eventually he began posting the tutorials to YouTube and was surprised by how popular they were. 

A multi-degree holder from MIT and Harvard, Khan soon discovered that this type of teaching was his passion and formed his Khan Academy site. Interest in, and praise for his site has been remarkable. He quit his day job as an investment analyst and devoted himself full time to producing his tutorials. Many of these videos were produced in a small walk-in closet in his home—the only space he could find where he could work relatively uninterrupted. 

Thanks to some big infusions of cash from the Bill and Melinda Gates Foundation and Google, Khan now has an office and employees. His software engineers are creating a management system to advance students appropriately though his instructional videos. 

A recent article in Wired magazine describes Grade 5 students in California successfully using Khan Academy and their management interface to learn math—some of these students had progressed to doing university level math! Nothing really new about Khan's concept: its mastery-based drill and practice design has been around since the beginning of personal computing. However, Khan's material is very different from the typical over-produced, glitzy (and sometimes boring) educational products on the market. He adds a very human touch, sounding like an enthusiastic, but very patient teacher as he delivers his lessons using fairly low tech means. He's also very good at explaining complicated topics in a simple way. 

Khan regulars say that they like that you can review the sections you don't understand over and over until you get it—an often stated virtue of technology based instruction.   Interestingly, it was an improvement to this aspect of our products that my customer was requesting a decade before. 

Khan Academy is already impacting classroom structure as more tweets and blogs focus on flipping the school day—having students learn at home and practice at school—but more on that topic in a future post.

Khan wants to start his own school. It will be interesting to track the success of this endeavor if it is realized. However, as with any instructional approach, I worry that too much learning with Khan's videos may become tiresome. I believe his products' strength may lie in a back up role supporting existing classroom instruction. 

If you haven't already, I suggest you have a look at Khan Academy. More importantly, suggest it to your students, other educators, your own children and their friends. Based on their comments, Khan's users seem to adore him. His website's motto, "Learn almost anything for free," is about as altruistic as possible. 

Pretesting: the Manageable Alternative to Retesting

Like many other teachers in the early 1990's, I was impressed by the promise of Outcome-Based Education or OBE. OBE was the new and improved version of mastery learning that stressed that all students can learn and that learning rather than content coverage should drive the pace of classroom instruction. Although I did, and still do, agree with many of the principles of OBE, I eventually became very disillusioned with the concept of student retesting.

Retesting is one of the key components of OBE. If students didn't master learning outcomes on a unit test, they were allowed to relearn those specific outcomes in a different way and then retest on those components until they had mastered them. This was where I found my first concern with retesting: the enormous amount of teacher time spent creating retests and relearning packages specifically tailored for individuals. My second concern was that my students weren't demonstrating the same gains in achievement as students in the literature were.

In fact, my students--as well-intentioned as most of them were--having diligently completed their relearning packages, showed up to the retest seemingly hoping that it was an easier version of the original. Generally, their results showed that they were not much better prepared for the retest than they were for the original test and very few significantly increased their score on the non-mastered concepts. I was demoralized and convinced that I was working harder at this than they were. As a result, over time, I largely abandoned the practice of retesting.

However, I wasn't quite ready to give up completely on OBE, I started offering pretests to students as opposed to retests.  I only had to create one really good pre-test per unit. Since I considered pretests as a type of formative assessment, they could be reused year after year and students could rewrite them as a review at any time. The critical components were that pretests be administered to students far enough in advance of the unit test to allow them time for reflection and review and that the students be provided with mastery reports by learning outcome based on their pretest performance.

The information contained in the mastery reports allowed the students to become metacognitive connoisseurs of their learning. Students began to come in for help much more often than they had prior to the implementation of pretesting, and accompanied by their mastery reports the students were much more aware of what they understood and didn't understand when asking for help.  Thanks to the information contained in the mastery reports, the students now knew where their learning gaps were. For example in biology class, students would point out that they had mastered the outcomes involving DNA replication but that they still needed some help with the translation steps of protein synthesis. These after school and in-class help sessions became very purposeful and focused.

The mastery reports produced a change in my instruction as well. Because the testing software could produce a mastery report summary by class, I could tailor my instructional and/or review activities to address the areas that most students were struggling with. The mastery reports took the guesswork out of lesson planning. I knew from these reports and my interactions with the students what they did and didn't understand and could revise my teaching accordingly.

Although pretests can be paper-based, posting them online increases student access and once again decreases teacher workload: no more standing in line at the photocopier and, depending on the type of question, no marking! Just make sure your online testing application is able to generate mastery reports for the students.

So what's the downside? It's time consuming to organize your test banks into valid questions grouped by learning outcome. However, whether you're pretesting or not, this level of organization helps you to create good tests quickly. And the opportunity to create and/or reorganize a test bank in such a way is a good collaborative activity for teachers in the same subject area. In order to create online pretests you need a sophisticated testing application. The one that I am most familiar with is LXR-Test. Although powerful, LXR is Windows only, not very user-friendly and expensive. Additionally, if you want online tests scored automatically, a feature that benefits students by providing instantaneous results, you are limited to multiple choice, matching and short answer questions. Although you may create and incorporate "essay" questions, they must be marked by hand and don't provide the instant results that are so beneficial to students.

I believe that pretesting students and providing them with mastery reports by learning outcome is a beneficial, practical, and efficient strategy that not only increases their achievement, but also helps them learn about learning. In addition the mastery reports also help teachers focus instructional time on the concepts that students need the most help with.