The Washington Post

Ambiguity, uncertainty, failure: Drivers for STEM Improvement

This was written by Arthur H. Camins, director of the Center for Innovation in Engineering and Science Education at the Stevens Institute of Technology in Hoboken, NJ.

By Arthur H. Camins

President Obama and countless reports all say that improving science and engineering literacy and ensuring a next generation of U.S. scientists and engineers are vital to our future. With the notable exceptions of creationists and climate change deniers, there is little opposition to making this an educational priority. However, current education policies at the state and federal levels contradict the core values of science and engineering, and are therefore likely to inhibit rather than catalyze progress.

Three values are at the heart of the practice of science and engineering and are central to discovery and innovation: searching for uncertainty, recognition of ambiguity and learning from failure. Therefore, it makes sense to nurture these values in school. In sharp contrast, popular education policies ignore ambiguities in assessment data and punish failure as determined by uncertain evidence. Unless checked, the latter will undermine the former.

Achieving a high level of scientific and engineering literacy means that students will have a lot to learn. They will need to know about how the natural world works, but also how to engage in the practices of scientific investigation that lead to new knowledge development. They will need to understand the creative design work of engineers and develop the skills to evaluate scientific claims and policy decisions related to the uses and consequences of technology. But, important as these are, they are not enough. Students also need to be fascinated and inspired, so that that they all value the work of- and some will want to become- scientists and engineers.

In his terrific recent book, “Ignorance: How it Drives Science,” Stuart Firestein eloquently makes the point that scientists are inspired to do their work not by demonstrating what they already know, but rather by exploring the mysteries of what they do not know. To similarly engage young people, we need to inspire them to want to investigate and design. Whether to spur interest in careers or to promote the values of science for citizen engagement, students will need a compelling reason to engage and work hard. They will need to see science and engineering as vehicles to make a difference in their world. In sharp contrast, the current over-emphasis on consequential assessment suggests to too many teachers and students that the goal of learning is correct answers and short-term test performance.

Similarly, engineers value what they can learn from their design shortcomings and failures. Duke University civil engineering professor, Henry Petroski, writes,

I believe that the concept of failure…is central to understanding engineering, for engineering design has as its first and foremost objective the obviation of failure. Thus the colossal failures that do occur are ultimately failures of design, but the lessons learned from these disasters can do more to advance engineering knowledge than all the successful machines and structures in the world.

The current testing regimen rewards what students know, stigmatizes what they do not know, and encourages hiding failure. Alternatively, the practice of science and engineering rewards what is discovered, what is invented and what is improved. The latter are not independent of scientific knowledge — explanations of how the natural world works. However, what is known changes as often unexpected, evidence sheds fresh light on old facts and new explanations emerge. The Framework for K-12 Science Education and its derivative Next Generation Science Standards go to great lengths to emphasize the interdependence of core ideas and practices.

However, putting this into classrooms will take time and explicit educational effort. It means focusing on some topics to the exclusion of others, because there is a direct relationship between knowledge depth and the ability to recognize uncertainty and ambiguity. It means building exploration of uncertainty into science curricula and making sure that engineering curricula includes lessons in learning from design failure . Curricular examples already exist, but for teachers to invest the required time, the emphasis in assessment will need shift away from information recall on summative tests to diagnostic and formative use of daily student work related to their ability to engage in the full range of science and engineering practices.

I’ve learned in a lifetime of home improvement work that the most important thing to know is what I do not know. I make mistakes when I act based on thin knowledge and unrecognized ignorance. So too, in almost 40 years as an educator, the more I know, the more I have appreciated the complexity of teaching and the many areas in which we are still ignorant. We make mistakes in education policy when we see certainty when there is ambiguity. We deprive students of the opportunity to learn and be inspired when we fail to provide them with sufficient time to explore their uncertainties.

If we want students to want to become — or at least appreciate the work of — scientists and engineers, then this is an essential intellectual disposition to develop early in their education. Unfortunately, current education policies that rely on uncertain data to reward success and punish failure inhibit achieving these goals in two significant ways.

First, they tend to cause schools to focus on short-term performance and diminish the time and focus required to support deeper student learning and interest. Second, they legitimize behaviors that undermine the very core values of science and engineering by punishing ignorance — not knowing — rather then using it as an inspiration for learning and improvement.

Unfortunately, ignorance has become a pejorative term. However, not knowing is simply a human condition in a complex, difficult to comprehend world. Therefore, ignorance is excusable and nothing about which anyone should be ashamed, unless, of course it is willful in the face of evidence or need. In fact, the inclination to search for and recognize areas of ignorance- areas of uncertainty and doubt- is a hallmark of science. The only shameful thing is to ignore uncertainly.

This happens in classrooms when unrealistic content coverage demands cause teachers to cut short exploration of variation in investigation data for what could be learned about the sources of measurement uncertainty, or when assessments are returned to students without spending time to address learning gaps. Uncertainty is the first and necessary step for insight, new discovery and innovation. While failure to solve complex problems is the natural result of necessarily limited knowledge, it can often be the unfortunate result of ignoring known uncertainty. While learning from failure is expected of scientists and engineers, ignoring uncertainty and ambiguity or punishing failure is not. And this, arguably, is an all too recognizable result of the current overreliance on limited and flawed assessment data for consequential decisions. It causes people to be cynical about the central idea of supporting claims with evidence, which in turn weakens the values of science and engineering.

A key feature of current policy is rating individual teachers with respect to their presumed effectiveness based on changes in the test scores of their students. While, many assessment experts have pointed to the uncertainties and ambiguities in data interpretation when used for this purpose and have challenged the veracity of claims relative to teacher effectiveness measures, policy makers have ignored these uncertainties. Is this willful ignorance? What is the inherent message of this policy to students and teachers about the integrity of the scientific practices of the adults who wield power? Researchers have pointed to the instability of teachers’ value-added scores across multiple years, but advocates for their use have dismissed this concern. Is dismissing high variation in results a good lesson for the next generation of scientists and engineers whose integrity we need to trust when they investigate drug interactions or design new airplanes. Do we want to students to learn to adopt such a “good enough” attitude for consequential decisions?

Many of the calls for improvement in scientific literacy reference a competitive disadvantage of U.S. students relative to their peers in Finland, Singapore and Canada as revealed on international assessments. While reforms in the United States stress market- and privatization-driven strategies, none of these countries have made advancement in this way. Notably, they have invested heavily in increasing the quality of in-depth teacher development and apprenticeship, raised the level of respect for the teaching profession and increased dedicated time for professional learning. Is moving ahead with a rather different, but predetermined, course in the face of contraindicated or absent data an example to students of effective science and engineering practices or is it a willfully ignorant bad example?

Science and engineering are advanced by a sense of wonder, a willingness to make tentative conjectures and the freedom to be wrong. What do students learn when they know that their scores on a test will result in reward or penalty for their teachers? Will this breed integrity or cynicism? Will teachers whose career advancement is predicated on student test scores emphasize the value of uncertainty, tolerance for ambiguity and learning from failure? Surely students will see the hypocrisy of an implicit message that says, “You need to take risks and learn from your mistakes so that you do well on tests and I won’t be fired.”

Another important feature of the practices of scientists and engineers and of the modern workplace is collaboration across disparate cultures and expertise. In fact, the resulting interchange and critique are often the source for pointing out uncertainty and spurring innovation and discovery. Students develop values and behaviors by observing the behavior of the adults around them. Evaluation schemes being promoted by federal and state policies require school districts to rank order teacher effectiveness on a normative scale. In other words, no matter what, some teachers must be at the top of the rating scale while others will be at the bottom. As a result, teachers recognize that personal advancement demands being rated better than colleagues. Will this policy promote a teacher demonstration of collaborative knowledge sharing for a common goal or competition for individual reward?

Do policy makers want to know which adult behaviors engender in students the talents and values required to practice science and engineering? Or is the pursuit of policies that do the opposite, willful ignorance?


Follow The Answer Sheet every day by bookmarking .

Show Comments
Washington Post Subscriptions

Get 2 months of digital access to The Washington Post for just 99¢.

A limited time offer for Apple Pay users.

Buy with
Cancel anytime

$9.99/month after the two month trial period. Sales tax may apply.
By subscribing you agree to our Terms of Service, Digital Products Terms of Sale & Privacy Policy.

Get 2 months of digital access to The Washington Post for just 99¢.

Most Read
Read stories based on reporting for “Trump Revealed,” a broad, comprehensive biography of the life of the president-elect.


Success! Check your inbox for details.

See all newsletters

Close video player
Now Playing