Skip main navigation.
U.S. Department of Education: Promoting Educational Excellence for all Americans - Link to ED.gov Home Page

Research on Science Education

 
Text for slide is below.
previous arrow Previous Table of Contents Next next arrow

Slide 18

Discovery learning

  • Piagetian conservation learning
  • LOGO programming
  • Misconceptions

Speaker's notes:

Discovery learning is a prime example of a questionable extrapolation from a science of learning. The evidence is now overwhelming that students do not develop superior understanding of scientific content when they are encouraged to figure things out for themselves, and the evidence goes way back.

The Swiss psychologist, Jean Piaget, posited that children have to learn on their own, and that the best circumstance for learning is one in which the child’s existing way of thinking is confronted by observations that don’t fit. This forces the child to accommodate to this new experience, which produces cognitive growth. Piaget’s favorite task involved conservation. Suppose a child is confronted with two identical tall beakers of water. In front of the child’s eyes the content of one beaker is poured into a low wide container. The child is asked to whether the two containers, the tall one and the low one, contain the same amount of water. Most children prior to middle elementary school will say that the tall one has more water.

Hundreds of studies were done on teaching children to understand that liquid volume and other properties of matter are conserved despite changes in appearance. Piaget dismissively called it the “American question.” But those studies showed clearly that direct instruction to children in the form of explanations, corrections, and rules could successfully teach children to answer conservation problems correctly, when no amount of discovery learning worked. Further, the learning produced by direct instruction stuck in that children still answered correctly weeks or months later, and they were not easily talked out of their answers.

A similar finding emerged in the later research on children’s learning of programming skills in the graphical LOGO language. The idea, inspired by Piaget, was that children would develop superior understanding if they worked on their own with LOGO. However, numerous studies showed that children given specific instructions on programming LOGO learned more. For example, Fay & Mayer taught LOGO using either pure discovery or by giving students the same projects along with explicit instruction on design concepts such as modularization, and with hints and feedback. On subsequent tasks, the guided group wrote more elegant programs, and solved planning tasks much better than discovery group.

Richard E. Mayer (2004). Should there be a three-strikes rule against pure discovery learning? American Psychologist, 59(12), 14-19.

Feedback, modeling, explanations all work compared to discovering the principle

Fay & Mayer taught LOGO using either pure discovery or guided discovery (students given the same projects along with explicit instruction on design concepts such as modularization, hints and feedback). On subsequent tasks, guided discovery group wrote more elegant programs, and solved planning tasks better than discovery group.