Skip to Navigation
University of Pittsburgh
Print This Page Print this pages

December 9, 2010

What kind of instruction puts students on the path to expertise?

Cognitive research into education often focuses on the habits of experts as opposed to the behavior of novices. However, it takes more than understanding efficient problem solving to put students on the path to expertise, said a Stanford education professor in last month’s Teaching Excellence Lecture.

What sort of instruction puts students on a trajectory toward expertise? And, how do we know instructors are succeeding? That depends on the type of expertise called for, said Daniel Schwartz, co-director of the multi-institution Learning in Informal and Formal Environments National Science Foundation Science of Learning Center, in his talk, “Trajectories of Efficiency and Innovation in Teaching and Learning,” hosted by the School of Arts and Sciences.

Routine vs. adaptive expertise

Two dimensions of learning are needed, Schwartz said: Students need help initially to progress from novices with routine expertise to novices with adaptive expertise — with the further goal of putting them on the path to being an adaptive expert high both in innovation as well as efficiency.

“Adaptive expertise is probably the better goal for college education given that jobs and knowledge based future will continually change. You need to be prepared to adapt when you go on to the informal space of the workplace,” Schwartz noted.

Routine expertise — defined as a high level of efficiency at performing a recurring task — depends on a stable environment, but also has a role, Schwartz said. “We can’t just teach them efficiency. We need to prepare them to adapt and learn new solutions when they get out there.”

Innovation and efficiency may be viewed as opposites — such as the creative person and the drudge, Schwartz said. “That’s just not true. We need both.”

Innovation has its roots in routine expertise, he said. Part of being able to come up with an innovative concept is based in mastering a sufficient body of knowledge. Progressing in innovation, however, requires a bent for exploration.

Gaining expertise

One aspect of expertise lies in becoming increasingly efficient, something that is important for routine tasks, Schwartz said. He gave the example of learning to make a left turn while driving a stick shift. The task is difficult for novices, but experts can perform the maneuver with ease.

A great deal is known about improving efficiency, he noted. Memorization, practice through problem solving and applying knowledge by practicing in contexts similar to real-world situations all are helpful. Feedback also is important to help novices improve their performance.

Educated perception is another mark of expertise. Experts see in a way that is more precise than what novices observe, Schwartz said. “Experts see precise structures in situations. They go beyond the basic level of interpretation.” Displaying a photo of a bird, he elaborated: “To me this is a small blue bird. To an expert, it’s an indigo bunting.”

Another aspect of expertise is the ability to see structure amid variation. “It’s not just noticing features, it’s noticing structures across many features,” he said.

Contrasting cases can help students move beyond the basics, Schwartz said, offering as an example tasting wines side-by-side to discern their characteristics. Contrasts in the form of negative instances also are helpful. “A lot of times we think the best way to teach is to just show [students] the right answer. Sometimes showing near-misses helps a lot. The contrast can help isolate the feature of interest,” Schwartz said, noting that anything a student sees can contain infinite information. “The trick is figuring out which thing is important.”

Efficiency vs. innovation

Efficiency-first instruction leads to routine expertise and is a good approach to instruction for recurrent tasks in stable environments, he said.

However, this sort of instruction can come at the expense of being adaptive, he said, proposing innovation-first instruction for knowledge that will require adaptive expertise.

In an efficiency-first setting, students tend to pay attention to the facts or formulas being presented. Schwartz said, “The result is they don’t go beyond the basic level of perception. Additionally, because you’ve given them the solution, they’ll just solve each problem one at a time with that solution. They won’t see the structure. The consequence is they fail to transfer to new situations.”

Adaptive expertise

Adaptive expertise requires new ways of seeing and understanding, Schwartz said. “You’ve got to continually try to come up with a new structure.”

While instructing students in contrasting cases can get them beyond the basics, “You also need to push students to seek a general explanation,” he said. “You want them to innovate a general explanation that handles all the cases — even the cases they haven’t seen yet.”

For example, a study of students from information-intensive disciplines — computer science, biology, psychology and engineering — compared the way undergraduates differed from graduate students when presented with a problem-solving task.

Students received outlines of 12 medical cases — each a single sheet that described the symptoms and diagnosis. They were to diagnose 10 new patients whose cases appeared on a computer screen. They selected tests they wanted to run, then entered a diagnosis.

Although graduates and undergraduates were equally accurate, the undergrads nearly had finished before the grad students diagnosed even one case.

Analysis showed that graduates created a visual representation  — making a matrix out of the pages. They had representations to cover all possible representations they might see rather than handling one at a time, as most undergrads did. The undergraduates tended to solve each problem individually, shuffling through the sheets of paper case by case instead of modeling the whole situation.

The graduates showed a different sort of expertise that was not merely efficiency driven, Schwartz said. Such behavior demonstrates prospective adaptation. “They changed the task ahead of time by introducing a representation to help them solve the group. Their expertise allows them to adapt the situation ahead of time.”

Pushing students to seek a general explanation that handles all cases — even the ones they haven’t seen yet — can help them gain adaptive expertise, he said.

Schwartz cited an experiment in which students were shown several simulations in which a magnet, when moved in various ways around a light bulb, turned it on or off.

One group was pressed to find a single explanation to cover all the cases, while the other was instructed to predict, observe and explain each instance.

Those who were pushed to find the “big story” behind the first three cases — an inductive approach — did four times better at explaining a subsequent case than those who were led in the classic hypothetic approach.

The way to adaptive expertise

When it comes to traditional curriculum sequences, many scientific disciplines are heavy on math and analytical courses early on, leaving opportunities for innovation until much later. Schwartz questioned that approach. “Should we tell them what they need to know first, then let them go innovate?” he said. “The answer, I think, is no.”

In a study of 8th graders, one group was taught about density using a standard tell-and-practice approach. Students got an explanation of the concept and the formula, then were shown solutions to sample density problems before being asked to solve more. Another group of students was instructed to invent a “crowdedness index” without being taught about density first.

Each group then was split into subgroups that received slightly different worksheets. Some got an abstract version. Others got illustrations of dots drawn inside cubes to represent varying levels of density. Some got an illustration of clowns on a bus with crowdedness representing density.

When later asked to re-draw the worksheet from memory, many students missed the concept.

“The students who received the seemingly efficient instruction missed the basic structure of the phenomenon that density was a ratio,” he said.

When tested later on their ability to use the concept to explain the stretchiness of trampoline fabric, students who were asked to invent the crowdedness index were four times more likely to use a ratio. Others who had been told density is a ratio and were shown the ratio to use, failed to transfer the idea to the new case.

“This is why I don’t like doing the efficiency first,” Schwartz said. “It’s going to lead to routine expertise. They’re just going to apply what you tell them and they’re not going to be ready to adapt to a new situation.”

Innovation vs. efficiency

Does the innovation-first method come at the expense of efficiency?

Efficiency-oriented instruction can block subsequent adaptiveness if students pay attention to solutions instead of problems, but innovation-oriented instruction does not block subsequent efficiency, Schwartz said. Efforts to generate a general explanation prepare students to learn the efficient solution more quickly.

It didn’t matter that some students in the innovation group struggled but didn’t come up with the answer, Schwartz said. In the end, students were given the density formula and everyone practiced several problems. “Giving them this chance to innovate doesn’t hurt their subsequent efficiency if you then tell them the efficient solution,” Schwartz said.

On the path to adaptive expertise

How can students be put onto the trajectory toward adaptive expertise?

Prepare them for learning first, Schwartz said. Experts often learn well from lectures because they have sufficient prior knowledge, he pointed out. Novices, because they often lack that knowledge, can be harder to engage.

Blaming previous teachers and lecturing more aren’t the answer, he said. “A different way to think about it is to create ‘a time for telling.’ See if you can prepare them so they can make sense of your lecture.”

In one experiment that tested whether students were prepared to learn from a lecture, cognitive psychology students received a homework assignment in which one-third of the group was asked to summarize a book chapter that described classic experiments using graphs, results and the relevant theories. The other two-thirds were assigned to graph the important patterns from simplified descriptions and data, but weren’t told what was important.

Schwartz noted that the graphing activity featured the concept of contrasting cases — “because experiments are contrasts” — and innovation because students were asked to notice the differences and make a general formulation on their own.

Later, half of the graphing group, as a control, did additional graphing. The other students attended a lecture that reviewed the experiments, their findings and the theories.

A week later, students were assigned to predict the outcome of a new study on memory. An assessment (based on whether they found eight predictors relevant to the lecture) showed that the group that did the summary followed by the lecture did worst. The group that did graphing and more graphing fared better, but those who did the graphing then had the lecture performed best.

Graphing the data raised students’ innovation, while the lecture nudged them toward further efficiency. “Then the new situation extends it so they’re starting to think about it in a new context,” he said.

The graphing activity prepared students to learn from the lecture, he said. Innovation activities create a time for telling, enabling students to “get” what theories are designed to explain.

“Instead of just telling them the answers, they understand what it is that it’s an answer to,” Schwartz said. “The benefits to this are probably going to appear when you assess their abilities to adapt understanding to handle new situations.”

Assessment

How can we know which trajectory students are on? Are they moving toward adaptive expertise or becoming routine experts? Perhaps the tests need to change, Schwartz said. “The problem is that our current assessments are actually measuring routine expertise, not adaptive expertise.”

Most often, testing is in the form of sequestered problem solving (SPS) — the student with nothing but a No. 2 pencil and a bubble form. “This is a great way to test efficient knowledge: How well can you just retrieve and apply this under circumstances without any support?” he said.

“This dominates education and feeds into the efficiency mentality,” Schwartz said, arguing for the elimination of that testing environment.

For example, he described two fictional job applicants: Bob, who took a five-week course in Excel, and Mike, who doesn’t know Excel, but taught himself three other spreadsheet programs.

In a job interview with a company that uses Excel, Bob is likely to come out ahead if given an SPS test filled with questions on Excel. But in the long run Mike, who understands the underlying principles of spreadsheet software, may be the better candidate.

Schwartz said a different kind of assessment — one of preparation for future learning (PFL) in which students are given resources and tested on what they can learn — would be better for discovering what students can learn.

“Sequestered problem solving measures routine expertise, efficient retrieval of information. The PFL measures are the ones measuring the trajectory of adaptive expertise,” he said.

—Kimberly K. Barlow

Filed under: Feature,Volume 43 Issue 8

Leave a Reply