American Scientist

Anyone who has seen a toddler pick up a smartphone and immediately, seemingly intuitively, know how to use it has witnessed the ways in which computers have changed how we think. As algorithms have become increasingly enmeshed in everyday life, we have adapted our thought processes to better utilize and communicate with them. In their book Computational Thinking, Peter J. Denning and Matti Tedre examine the development of these thought processes and show how they help us understand everything from software development to biology. The authors define computational thinking as “the mental skills and practices for designing computations that get computers to do jobs for us, and for explaining and interpreting the world as a complex of information processes.” Many aspects of computational thinking have a long history that predates the invention of electronic computers. Now curricula aimed at preparing students for the modern world are attempting to teach computational thinking. In this excerpt, Denning and Tedre recount the process of integrating computational thinking into the education of students in kindergarten through grade 12.

Through the 1990s, computational thinking (CT) education was mostly the purview of universities; very little CT education was available elsewhere. Pre-college K–12 schools had a scattering of computer courses; most focused on computer literacy and a handful on programming. A tipping point came after 2000, when many people saw how pervasive computing was in everyday work and home life. Educators and policymakers began to agree that understanding the mechanisms of digitalization is an important 21st-century skill.

The previously obscure notion of the algorithm entered everyday conversation as people cited value they had received from algorithms on their web searches, income tax preparation, online shopping, spreadsheets, neatly formatted documents, display-ready presentations, and computerized courses, and then later on smartphones, social networks, ride hailing, short-term renting, dating, finding friends, and much more. It seemed that understanding how it all worked was central for coping in the modern world. It was finally time to bring computing to the K–12 level of education.

CT Is Not Easily Transferable

The first wave of bringing CT to K–12 schools focused on programming. In the mid-1960s, some U.S. high schools got DEC PDP-8 minicomputers, and enterprising teachers organized courses around them. Numerous initiatives for using computers in schools over the 1960s and 1970s led to a few notable innovations. For example, the Little Man Computer for teaching machine languages and computers to students was introduced in 1965; one of the early programming languages for children, Logo, was introduced in 1967; and the famous concept for Dynabook, a children’s portable computer, was born in 1968. Although minicomputers and some microcomputers were common in the late 1970s, educators, lacking financial resources and political will, were unable to transform the pilot courses into a large-scale rollout to schools.

The Logo programming language was a standout among the many initiatives of the 1960s. It was a part of an integrated framework of pedagogical, technological, and educational ideas designed by Seymour Papert, based in his deeply grounded understanding of how children learn. His 1980 book Mindstorms, written after a decade of research and experimentation with Logo, was a milestone for computing education and for teaching computational thinking.

Papert coined the phrase computational thinking for the practice of procedural thinking he taught to children. He argued that learning is most effective when learners “construct knowledge” —they build their knowledge from practicing it rather than being told. The learning theory of constructionism became very popular in education. Papert continued to advocate self-directed learning, project learning, meaningful representations, facilitation-based education, and the use of technology to support learning in the classroom. But the central idea of Mindstorms—the shift from “learning to program” to “programming to learn”—remained hard to market to teachers.

The hope that a small number of teachers could teach CT to everybody was paired with the transfer hypothesis—a belief that CT is a metacognitive skill learned from programming, and that students who learn CT in one domain become better problem-solvers in other domains, too. This belief bolstered the position that teaching computing should be an essential element of K–12 education. Supporters argued that learning programming improves generic thinking skills such as logical thinking and generally “sharpens the mind.”

Critics of the transfer hypothesis referred to a research base in developmental cognitive science, arguing that there was no evidence of skill-transfer from programming to other subjects. Research with adults did not support transfer of cognitive skills between domains. Programming itself is a complex network of skills including mathematical abilities, conditional reasoning, analogical reasoning, procedural thinking, temporal reasoning, and memory capacity. It was not clear which parts of this complex transferred and which did not. After much detailed investigation, education researchers eventually concluded there is not enough evidence to accept the transfer hypothesis. It was not compelling as a justification for teaching computing in K–12 schools.

From Literacy to Fluency

The early advocates of algorithmic thinking would be appalled at many of the “computer literacy” courses in the 1980s and 1990s, which focused on how to use desktop applications, such as word processors, spreadsheets, and sketchpads. Literacy with desktop software was a far cry from their aspirations to participate in and shape the computer revolution. Professional societies offered to help K–12 educators develop courses with more depth, but got little buy-in. In 1999, a U.S. National Research Council (NRC) commission upped the ante, reframing the question from literacy to fluency. Fluency offered capabilities, concepts, and skills essential for some levels of computational thinking. The NRC initiative was paired with a textbook titled Fluency with Information Technology, which became quite popular among high school teachers.

Educators and parents wanted to ensure that students were learning 21st-century skills that prepared them for employment in STEM fields.

• Facebook
• Twitter

Many schools brought computing into their curricula for pragmatic reasons as they responded to demands from parents and school boards. They wanted to ensure that students were learning 21st-century skills that prepared them for employment in STEM fields, broadened social participation, and provided a new means to express individual creativity. Educators and parents were favorably disposed toward these goals because they believed that learning programming teaches important skills no other subject does, and because they did not want their children to be at a disadvantage in a world increasingly dependent on skills with information and communication technology.

In the 2000s, the entry of programming and computational design into schools was also easier because of advances in programming methodology and technology, and changes in what entry-level programmers needed to know. New languages such as Python were much easier to use. Graphical, drag-and-drop user interfaces were very successful. Powerful tools automated significant parts of the programming process. With all these advancements in languages, tools, and methods, programming was accessible to more students and teachers than ever before. But even so, in 2010 many schools had no computer courses or Advanced Placement curriculum in computing.

Computational Thinking Revived

In 2006, Jeannette Wing, then an assistant director at the U.S. National Science Foundation (NSF), published an essay in Communications of the ACM proposing that computational thinking is what everyone wants—not literacy or fluency. The essay launched a new wave in the movement to provide computing courses for all students in K–12 schools. The term computational thinking resonated and inspired action where literacy and fluency had not. Wing mobilized significant resources at the NSF to bring researchers into investigations of CT in education, to train teachers for teaching CT, to mobilize private organizations to produce K–12 curriculum recommendations for CT, and to develop a new Advanced Placement curriculum and exam on computing principles. Wing’s essay became one of the most cited in computing education, a rallying point in a global movement to promote the penetration of CT into K–12 education.

Major organizations developed and recommended curriculum frameworks for K–12 CT. These organizations promoted coding clubs, coding boot camps, and the international movement called “Hour of Code.” CT became a keyword gathering hundreds of thousands of hits in news stories, blog postings, book chapters, articles, research projects, and essays on computing education.

From Computational Thinking. Copyright © 2019 by the Massachusetts Institute of Technology. All rights reserved. Forthcoming from MIT Press May 2019.