Ad Code

Know Constructivism in Computer Science Education

Constructivism in Computer Science Education

Hi Friends,

Studies have shown that relatively few students reach an acceptable level of achievement in high-school science and mathematics (Duit, 1991). Physics teachers seem to have the worst time, as students retain a naive the-ory of physics despite intensive instruction in Newtonian mechanics (McCloskey, 1983).


For constructivists this is not surprising: everyone who has ever thrown a ball—that is, everyone—knows that if you don’t keep applying force, an object in motion will eventually come to rest. Apparently, these ideas are so entrenched that mere lectures and even experiments have a difficult time evicting them. At most, a certain facility in manipulating formulas is achieved, but this fails as soon as the student attempts to solve a problem that requires deep understanding. 

The discrepancy between performance and understanding has also been noted in mathematics education: The pupil’s fundamental problems with such ideas as negative or complex numbers tend to be overlooked by the teacher mainly because the latter’s own implicit beliefs make him or her oblivious to the possibility of somebody having a different ontological stance....Another circumstance that helps in concealing ontological difficulties is the fact that a student may become quite skilful in manipulating concepts even without reifying them. (Sfard, 1994, p. 268) Physics educators are very receptive to constructivist principles.  

After all, physicists have undergone two massive restructurings of their world within a short period of history: from Aristotelian physics to Newtonian physics and then to Einsteinian physics. One cannot fault them for their reluctance to believe that E=mc2 is an absolute truth. This openness is demonstrated by their willingness to attribute to the student alternative frameworks rather than misconceptions. In fact, von Glasersfeld, a pioneer of constructivism, would never say that something is wrong, because he does not believe in the possibility of establishing universal truths.  

Instead, he says that concepts are viable “if they prove adequate in the contexts in which they were created’’ (Glasersfeld, 1995, p. 7). This is analogous to the use of the word in biology to denote an organism adapted to its environment. The box metaphor for variables, and the communications model of reference parameters (discussed later) are simply nonviable, because they cause the student to fail on programming tasks. 

According to constructivism, a teacher cannot ignore the student’s existing knowledge; instead, he or she must question the student in order to understand exactly what theory the student is currently using, and only then attempt to guide the student to the “correct” theory. It is perhaps axiomatic for a constructivist that students have consistent theories—they just happen to be at variance with the (currently accepted) scientific theory. In most fields of science education including computer science, there is a large body of research that catalogs misconceptions. 

A constructivist would view a misconception not as a mistake, but as a logical construction based on a consistent, though nonstandard theory, held by the student. Even Matthews—who is critical of constructivism—is careful to point out that: 

“It is with respect to [contemporary physics] that [students] have misconceptions, it is not with respect to the behavior of the natural world” 

(Matthews, 1994, p. 133). Merely listing misconceptions is fruitless; a misconception must be accompanied by a description of the underlying model that caused it, and by a suggestion how to base the construction of a viable model on the existing one. Smith III, dissed, and Rochelle (1993) go so far as to claim that misconceptions form the prior knowledge that is essential to the construction of new knowledge! It is important not to confuse the use of computers in science education with the study of computer science. Computers are often seen as a tool to increase the constructive content of science education.  For example, Hatfield (1991) considers programming, or more generally algorithmics, as constructive. However, his article is essentially concerned with the contribution of algorithmics to mathematical education, rather than to the constructivist aspects of computer science and programming. 


“The role of the computer activities provide an experiential basis for all other learning modes....the main point is spending the time and effort on the problem, not solving it” 

(Leron & Dubinsky, 1995, p. 231, 236). In CSE, the computer is not just providing an experiential basis, nor is it creating a microworld (Harel & Papert, 1991) in order to facilitate construction of knowledge in another domain. 

Instead, the students are learning about computing itself—systems, algorithms, languages—and lessons from the use of computers in other fields must be applied carefully.

Post a Comment