Discovery learning refers to various instructional design models that engages students in learning through discovery. Usually the pedagogical aims are threefold: (1) Promote "deep" learning, (2) Promote meta-cognitive skills (develop problem-solving skills, creativity, etc.), (3) Promote student engagement.
According to van Joolingen (1999:385): “Discovery learning is a type of learning where learners construct their own knowledge by experimenting with a domain, and inferring rules from the results of these experiments. The basic idea of this kind of learning is that because learners can design their own experiments in the domain and infer the rules of the domain themselves they are actually constructing their knowledge. Because of these constructive activities, it is assumed they will understand the domain at a higher level than when the necessary information is just presented by a teacher or an expository learning environment.”
According to Borthick & Jones (2000:181): “In discovery learning, participants learn to recognize a problem, characterize what a solution would look like, search for relevant information, develop a solution strategy, and execute the chosen strategy. In collaborative discovery learning, participants, immersed in a community of practice, solve problems together.”
According to Judith Conway's Educational Technology's Effect on Models of Instruction: “Jerome Bruner was influential in defining Discovery Learning. It uses Cognitive psychology as a base. Discovery learning is "an approach to instruction through which students interact with their environment-by exploring and manipulating objects, wrestling with questions and controversies, or performing experiments" (Ormrod, 1995, p. 442) The idea is that students are more likely to remember concepts they discover on their own. Teachers have found that discovery learning is most successful when students have prerequisite knowledge and undergo some structured experiences.” (Roblyer, Edwards, and Havriluk, 1997, p 68).
Discovery Learning provides students with opportunities to develop hypotheses to answer questions and can contribute to the development of a lifelong love of learning. Students propose issues or problems, gather data and observations to develop hypotheses, confirm or refine their hypotheses, and explain or prove their problems. Apple Teaching Methods, Discovery Learning
Discovery learning is based on this "Aha!" method. 
2 Theory and models of discovery learning
Discovery learning can be traced back to authors like Rousseau, Pestalozzi and Dewey. In particular Dewey's emphasis on "experience" is in vogue again.
Modern discovery learning approaches relate to constructivist theory and therefore Bruner is considered a father of discovery learning by many authors. E.g. in the Encyclopedia of Educational technology one can find the following quote from Bruner “"Emphasis on discovery in learning has precisely the effect on the learner of leading him to be a constructionist, to organize what he is encountering in a manner not only designed to discover regularity and relatedness, but also to avoid the kind of information drift that fails to keep account of the uses to which information might have to be put."”(Bruner, 1962).
Another strong influence for some kinds of discovery learning (see microworlds is Seymour Papert's constructionism. Donald Clark in his discovery learning page puts the following statement: “"You can't teach people everything they need to know. The best you can do is position them where they can find what they need to know when they need to know it." - Seymour Papert”
Discovery learning is also strongly tied to problem solving (or learning how to solve problems under a more meta-cognitive perspective): “"Learning theorists characterize learning to solve problems as discovery learning, in which participants learn to recognize a problem, characterize what a solution would look like, search for relevant information, develop a solution strategy, and execute the chosen strategy."” (Borthick & Jones, 2000:181)
Some authors point out that discovery learning may increase content relevance and student engagement (actually an argument that can be made for all sorts of project-oriented learning.
Discovery learning, like most constructivist instructional design models is not easy to implement, since learners need to possess a number of cognitive skills and be intrinsically motivated to learn.
van Joolingen (1999:386) makes the following point:
Therefore one must try to support discovery learning processes, however with the risk of disrupting the very nature process that should engage the learner in autonomous knowledge construction.
Of course, there is a lot of disruption of this "pure model". A lot of research has pointed to out that "unguided instruction" can fail to meet precise instructional goals. Therefore, in practice, most current forms of discovery learning are guided in various ways.
3 Models of discovery learning
- Collaborative discovery learning
- Discovery learning with microworlds
- Experiental learning (to some extent)
- Guided discovery learning
- Incidental learning
- Learning by exploring (exploratory learning)
- Simulation-based learning
- Case-based learning
- Problem-based learning
- inquiry-based learning
- Cognitive tools
- A simple combination of webpages (read/write) and forums or alternatively a Wiki
5 Advantages and disavantages of discovery learning
The discovery learning literature often claims the following advantages:
- Supports active engagement of the learner in the learning process
- Fosters curiosity
- Enables the development of life long learning skills
- Personalizes the learning experience
- Highly motivating as it allows individuals the opportunity to experiment and discover something for themselves
- Builds on learner's prior knowledge and understanding
- Develops a sense of independence and autonomy
- Make them responsible for their own mistakes and results
- Learning as most adults learn on the job and in real life situations
- A reason to record their procedure and discoveries - such as not repeating mistakes, a way to analyze what happened, and a way to record a victorious discovery
- Develops problem solving and creative skills
- Finds new and interesting avenues of information and learning - such as gravy made with too much cornstarch can become a molding medium
These sorts of arguments can be regrouped in two broad categories
- Development of meta cognitive skills (including some higher level cognitive strategies) useful in lifelong learning.
Most researchers would argue that pure discovery learning as a general and global teaching strategy for beginning and intermediary learners doesn't work. The debate on how much guiding is needed is somewhat open. See Kirschner et al. (2006) for a good overview (or Mayer, 2004; Feldon) and also Merrill's first principles of instruction model that does promote unguided problem-based learning at the final stages of an instructional design.
Typical criticisms are:
- (Sometimes huge) cognitive overload, potential to confuse the learner if no initial framework is available, etc.
- Measurable performance (compared to hard-core instructional designs) is worse for most learning situations.
- Creations of misconceptions ("knowing less after instruction")
- Weak students have a tendency to "fly under the radar" (Aleven et al. 2003) and teacher's fail to detect situations needing strong remediation or scaffolding.
- Some studies admit that strong students can benefit from weak treatments and others conclude that there is no difference, but more importantly they also conclude that weak students benefit strongly from strong treatments.
DSchneider thinks that despite very strong arguments (Kirschner et al., 2006) in disfavor of even guided discovery learning models like problem-based learning, the debate is still open. Most really serious studies concerned high-school science teaching. Now, science is very hard and indeed puts a very heavy load on short-term memory. In addition, in order to solve even moderatly complex problems a person must engage many schemas. If nothing is available in long term memory, the learner is stuck.
As an example, DSchneider (from his own experience) doesn't believe that object-oriented programming could be taught by a discovery approach. Making web pages on the other hand could. Students can incrementally work on their own projet and integrate independent concepts like HTML, CSS, Ergonomics, Style, Color etc. on their own pace. A project-oriented approach to web page making probably also would be less effective than a strategy like direct instruction. On the positive side, students engaged in discovery with some scaffolding and monitoring provided by the teacher will learn to find resources, to read technical texts found on the Internet, to adapt a solution to their skill level (learn something about the economics), to decompose a problem, etc. I.e. they learn some skill that are probably transferrable to similar autonmous learning situation (e.g. learning SVG on their own).
6 Planning a Discovery Learning Experience
(paste by Stek, from http://members.aol.com/kitecd2/artcl_disclearn.htm#PLANNING)
- select an activity. To begin pick an activity that is relatively short so that follow-up attempts are easier to predict and plan for. Select a subject with which you are personally familiar and comfortable. Also in the beginning it is often best to choose an activity that does not have just one correct answer. Role-playing, creating sculptures, observing characteristics of objects, or searching for or classifying similar items all work well.
- gather materials. Remember to have enough materials for each learner to repeat the activity at least once.
- stay focused. Avoid learning tangents that may be interesting but will keep the learner from finishing the project, unless they are truly of great curiosity and value. Instead take notes concerning the new interest to follow-up on once the initial activity is completed.
- use caution. While the idea of discovery learning is for the instructor to step back and observe allowing the child to work independently, be sure that safety is observed. Activities such as cooking and cutting should always be supervised by an adult and experimenting with magnets is nice unless an important video or cassette tape is ruined.
- plan extra time. Understand that children working on their own will most likely take longer than they would with an adult moving them from step to step. Also be sure to plan time for repeated activities in case there is a failure or other reason to repeat the activity.
- record process and results. Include in the activity a requirement for older children to record their procedure and results. For young children guide, assist, or model record keeping.
- discuss and review. After and activity is completed and before it is repeated a second time (if needed), discuss the activity and its outcome with the child. Use the records which were kept to assist during this step. Once the activity has been analyzed, record any observations or mistakes.
- try again. Have the child repeat the activity if necessary. Encourage her to take into account what was done and the discussion that occurred. Allow her to use any records that were kept to assist her in successfully completing the activity. Give assistance and guidance as necessary.
- plan for more discovery learning activities. Think over how this activity worked for the child. As you plan more discovery activities take the answers to these questions into consideration. What went well? What could have gone better? How can any problem areas be corrected or alleviated?
- See http://copland.udel.edu/~jconway/EDST666.htm#dislrn (to be inserted in simulations.)
- Aleven, V., Stahl, E., Schworm, S., Fischer, F., & Wallace, R. (2003). Help Seeking and Help Design in Interactive Learning Environments. Review of Educational Research, 73(7), 277-320.
- Borthick, A. Faye & Donald R. Jones (2000). The Motivation for Collaborative Discovery Learning Online and Its Application in an Information Systems Assurance Course, Issues in Accounting Education, 15 (2). .
- This article presents a case study in detail
- Bruner, J.S. (1967). On knowing: Essays for the left hand. Cambridge, Mass: Harvard University Press.
- Davidson, N. The Small-Group Discovery Method in Secondary- and College-Level Mathematics, In N. Davidson (cd.) Cooperative Learning in Mathematics: A Handbook for Teachers. Addison-Wesley (Menlo Park, Ca.). 1990.
- Dunbar, K. 1993. Concept discovery in a scientific domain. Cognitive Science 17: 397-434.
- Feldon, David F. Perspectives on Learner-Centeredness: A Critical Review of Definitions and Practice, University of Southern California Center for Learning, PDF, retrieved, 17:17, 15 September 2006 (MEST).
- Jong, T. de, & Joolingen, W.R. van (1998). Scientific discovery learning with computer simulations of conceptual domains. Review of Educational Research, 68, 179-202.
- Joolingen, Wouter van (1999), Cognitive tools for discovery learning, International Journal of Artificial Intelligence in Education, 10, 385-397 
- Joyce, B., M. Weil, and B. Showers. (1992) Models of Teaching. Allyn and Bacon (Boston, Ma).
- Kirschner, P., Sweller, J., & Clark, R. E. (in press). Why Unguided Learning Does Not Work: An Analysis of the Failure of Discovery Learning, Problem-Based Learning, Experiential Learning and Inquiry-Based Learning. Educational Psychologist.PDF Preprint
- Kirschner, P. A., Sweller, J., & Clark, R. E. (2006) Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educational Psychologist, 41(2), 75-86 PDF
- Klahr, D., & Dunbar, K. (1988). Dual space search during scientific reasoning. Cognitive Science, 12, 1-48.
- Mayer, R. (2004). Should there be a three-strikes rule against pure discovery learning? American Psychologist, 59(1):14-19 [Should there be a three-strikes rule against pure discovery learning PDF] (Access restricted).
- Njoo, M., & Jong, T. de (1993). Exploratory learning with a computer simulation for control theory: Learning processes and instructional support. Journal of Research in Science Teaching, 30, 821-844.
- Sweller, John & Graham A. Cooper (1985) The Use of Worked Examples as a Substitute for Problem Solving in Learning Algebra, Cognition and Instruction, 2:1, 59-89, DOI: 10.1207/s1532690xci0201_3
- Tobias, S, (1991). An examination of some issues in the constructivist-ISD controversy from an eclectic perspective. Educational Technology, 31 (9), 41-43.