Appendix 1 Some Learning Theory Background

In a behaviorist view, "Learning" can be defined as something that occurs when a learner acquires the capacity to do something. The LE designer must provide the conditions for this process. For each type of learning, some conditions work best and some don't. Let's look at 2 classifications:

Types of Learning

(according to Kearsley 1993):

Attitudes: "....Disposition or tendency to respond positively or negatively towards a certain thing (idea, object, person, situation)." Also: Choose to behave this or that way according to opinions and beliefs.
Factual Information (Memorization): Processing of factual information and remembering is tied to previous knowledge. Memory research has also a lot to say about processing constraints.
Concepts (Discrimination): Concept learning encompasses learning how to discriminate and categorize things (with critical attributes). It also involves recall of instances, integration of new examples and sub-categorization. Concept formation is not related to simple recall, it must be constructed.
Reasoning (Inference, Deduction): "Reasoning encompasses all thinking activities that involve making or testing inferences. This includes inductive reasoning (i.e., concept formation) and deductive reasoning (i.e., logical argument). Reasoning is also closely related to problem-solving and creative behaviors".
Procedure Learning: Procedures refer to being able to solve a certain task by applying a procedure. Once a procedure is mastered its excused usually does not take much effort (e.g. ftp a file). Cognitive theories like Act or Soar are interested in this, because procedures are important in diminishing cognitive load.
Problem-Solving: A good example is Newell & Simons information processing paradigm for the study of problem-solving and the concepts of "means-ends-analysis" and "problem space". According to their GPS framework, problem-solving involves the identification of subgoals and the use of methods (especially heuristics) to satisfy the subgoals. And important contribution was also the methodology of protocol analysis (of "thinking aloud methods" which has been extensively used by Anderson (87) to implement intelligent tutoring systems according to his Act* theory (Anderson 83).
Learning Strategies: can be learned too to some extent. Very much dependant on what you want to learn
Sensory-Motor:

Note that learning types can be strongly related to different kinds of cognitive task behaviors (that are being used while learning or that are targets for learning). As an example, Kearsley (93) lists the following types of task behaviors:

Searching for/receiving information (detects, observes, inspects, identifies, reads, surveys)
Processing information (categorizes, calculates, codes, itemizes, tabulates, translates)
Problem-solving (analyzes, formulates, estimates, plans)
Decision-making (examines, chooses, compares, evaluates)
Communication (advises, answers, directs, informs, instructs, requests, transmits)
Sensory-motor processes (activates, adjusts, connects, regulates, tracks)

By combining those two kinds of typologies one can imagine the "haystack" Instructional Design theory is faced with when trying to operationlize how to learn what.

Other categories of learning types has been proposed such as the ones by Gagné (Aronson 83:81, Gagné 87: 64), i.e. (1) Intellectual Skill, (2) Verbal Information, (3) Cognitive Strategy (problem solving), (4) Attitude, (5) Motor Skill. In any case, it think it is useful in this context to distinguish at the least the following basic categories:

Factual Information & Concepts (Verbal Information): Remember and discriminate things
Problem Solving & Reasoning (Cognitive Strategy): Apply general or domain-specific heuristics to problem situations
Procedural skills: Learn how to do simple or complex tasks more or less automatically.

Learning/Teaching Strategies &Principles

How can we have the learner use an appropriate learning strategy? In some learning environments (specially the fully computer-based ones) learning and teaching strategies are integrated into its design. In others they are delivered apart. Principles and Strategies vary according to the type of learning and different theoretical orientations.

Bruner (66), inspired by Piaget, focussed on how people construct new knowledge. His constructivist approach (discovery methods and intellectual stages) still inspires current theories.

Instruction must be concerned with the experiences and contexts that make the student willing and able to learn (readiness).
Instruction must be structured so that it can be easily grasped by the student (spiral organization of the curriculum).
Instruction should be designed to facilitate extrapolation and or fill in the gaps (going beyond the information given).

An other early contribution was Ausubel's (63) subsumption theory concerned with how individuals learn large amounts of meaningful material from verbal/textual presentations in a school setting (as opposed to rote or discovery learning). He initiated that instructional sequences should make content more meaningful for the learner. He postulates (cf. Kearsley 93, Reigeluth 83:339) that:

Instruction (of verbal information) should start with general knowledge that subsumes content presented by successive differentiation, i.e.the most general and ideas of a subject should be presented first and then progressively differentiated in terms of detail and specificity.
More generally, instructional materials should attempt to integrate new material with previously presented information through comparisons and cross-referencing of new and old ideas.

Both Reigluth's (83) "Elaboration Theory" and Merrill's (83) "Component Display Theory" are based on work by Bruner and Ausubel.

Other more recent lines of research combine cognitivist information theory with results from more traditional experimental memory research.

An example is the Act* Theory using Intelligent Tutors as a test bed (cf. Anderson 87). "According to ACT*, all knowledge begins as declarative information; procedural knowledge is learned by making inferences from already existing factual knowledge. ACT* involves three types of learning: generalization, in which productions become broader in their range of application, discrimination, in which productions become narrower in their range of application, and strengthening, in which some productions are applied more often. New productions are formed by the conjunction or disjunction of existing productions. It is interesting to compare these three types of learning with the three modes of learning (accretion, restructuring, tuning) proposed by Rumelhart & Norman (.)" (Kearsley: 93).

Principles:

Identify the goal structure of the problem space to the learner.
Provide instruction in the context of the problem-solving task.
Provide immediate feedback on errors.
Minimize the working memory load.
Adjust the "grain size" of instruction to account for the knowledge compilation process.
Enable the student to approach the target skill by successive approximation.

With partially automatized environments such as Hypertext course on the Web, the student should be told how to use the material, how to read it and what to do beside.

"A typical study skill program is SQ3R [applicable to concept learning/D.S] which suggests 5 steps: (1) survey the material to be learned, (2) develop questions about the material, (3) read the material, (4) recall the key ideas, and (5) review the material." (Kearsley: 93).

Table of Contents Next Chapter