With apologies for the pun (in honor of fall, I thought about doing something with school-leavers too, but spared you), an interesting article about the loss of science majors in higher education turned up in the New York Times recently. The article by Christopher Drew pointed out that America's challenges with educating in science and math are not confined to K-12, which draws much of the press and international comparisons. They point out that once in college, there's an enormous attrition from science and math majors as well.
Some colleges are successfully weaving more projects into the earlier college years. However, there may be even more benefit to learners by investing in restructuring courses to weave real-world tasks together with concept instruction.
The article identifies a number of possible problems: a lack of math preparation, the growing complexity of the topics, perhaps grade inflation making humanities courses more attractive, and the challenge that STEM courses often build directly on each other, so early difficulties cannot be left behind.
One other potential cause mentioned was that the design of college STEM instruction itself was flawed. Students are faced with especially challenging freshman year courses, often followed by a year or more of abstract courses, before getting any exposure to the real work of the domain. From a student quoted in the article:
“I was trying to memorize equations, and engineering’s all about the application, which they really didn’t teach too well,” [he said.] “It was just like, ‘Do these practice problems, then you’re on your own.’”
A number of colleges are trying to fix this. The article mentions Notre Dame as including design projects in the freshman year for engineers, and Worcester Polytechnic Institute includes time for research and social service projects in the last two years, along with optional first year projects.
“That kind of early engagement, and letting them see they can work on something that is interesting and important, is a big deal,” says Arthur C. Heinricher, the dean of undergraduate studies [at WPI]. “That hooks students.”
The article goes on to explain how students appreciate working closely with faculty members, the teamwork aspects, the confidence-building that comes from project work, and seeing practical impact early from their challenging studies.
However, there are other benefits not mentioned if you draw tasks from the domain deeper into the design of courses themselves: faster, better, more flexible learning.
David Merrill summarizes this well in his book chapter, First Principles of Instruction. Dr. Merrill, a leading instructional designer, has worked closely with learning scientists to turn laboratory results into systems of instructional design that can be applied at scale to real world learning.
Learning science shows that complex knowledge and tasks are most efficiently taught using worked examples with faded coaching. (See E-Learning and the Science of Instruction for more details.) This suggests designing a course of complex study as a series of increasingly complex tasks in a domain. The concepts and principles of the domain get taught to resolve new challenges as the tasks get more complex. A graphic from Dr. Merrill's chapter illustrates how task-centric instruction works:
The idea is to include much coaching and support early in the instructional program to help novices succeed at the simplest tasks in the domain. As the learner become more expert at increasingly complex tasks in the domain, the coaching is faded back; increasingly, the learner becomes experienced at tackling new complexities on their own.
A few keys to why this works so well:
This is not the same as some implementations of problem-based learning, where beginning students are thrown into the deep-end of complex problems. For students who already have some exposure (i.e., who are not really novices), rich, complex problem-solving is terrific – but for true novices in a domain, it can be highly confusing because they cannot “chunk” the information yet. That's why the first tasks for novices need to be simple, so that most can succeed.
Students are repeatedly doing full tasks within the domain from the beginning. A learner practices basic procedures and skills from the domain over and over, even as they expand the complexity of the concepts and work. This increases the automation of basic skills and tasks, which frees up working memory to work on the harder problems to come.
Working with tasks from the domain from the start keeps the interest of students. Instead of years of theory, each new conceptual challenge arrives in the middle of (and helps resolve) a domain task that they want to be doing professionally.
The approach instills a “habit of mind” in students to keep applying the principles and concepts of the domain (and looking for more) whenever they hit complexity on tasks. Rather than being “deer in the headlights” their first days of real work, they've already spent years tackling challenges in tasks by applying concepts and principles from the domain.
What's in the way of doing this? One problem is that it requires a great deal of thought up front to build a structure of simple-to-complex tasks that also follows the conceptual development of the domain. This is likely to be more than any individual faculty member can find time to do (esp. with the pressures of research), and so benefits from a coordinated team effort. Ideally, since STEM faculty usually have very little training in either learning science or instructional design, the team would include well-trained instructional designers experienced at this kind of thing, who could help draw out the real-world experience from faculty or others to structure this kind of course.
Of course, that increases the up-front cost of creating the learning environment. However, research shows that when you set up instruction this way, you improve both the retention and transfer of the skills to the new domain, and at the same time make instruction for each student take less time than more traditional approaches.
Hmm, this sounds suspiciously like what most industries have experienced over the last 100 years: intelligently invested capital up-front, helping us all to less costly, higher quality, services and products at scale.
Good heavens – is that sort of thinking allowed in our sector? ;-)
At Kaplan, we're starting to apply some of these ideas to our own learning environments, doing pilots to see if more investment up-front along these lines by a team can yield more compelling and effective courses for students down the road.
A long way to go, but looking forward to hearing more about these kinds of things from others as all of us work to systematically help students succeed at their dreams.