Effective Learning

learn more, learn better

User Tools

Site Tools


ref:koedinger2011knowledge

The Knowledge-Learning-Instruction (KLI) Framework: Bridging the Science-Practice Chasm to Enhance Robust Student Learning

Citation Koedinger, K. R., Corbett, A. T., & Perfetti, C. (2011). The Knowledge-Learning-Instruction (KLI) Framework: Bridging the Science-Practice Chasm to Enhance Robust Student Learning. Cognition. Sidewiki
BibDesk PDF

BibTex

BibTex

BibTex

@article{koedinger2011knowledge,
author = {Koedinger, Kenneth R and Corbett, Albert T and Perfetti, Charles},
date-added = {2011-09-10 17:52:58 -0400},
date-modified = {2012-05-29 23:00:46 -0400},
journal = {Cognition},
keywords = {cognitive load; memory; conceptual knowledge; representation; instruction; learning; knowledge components; spacing effect; practice},
read = {1},
title = {The Knowledge-Learning-Instruction (KLI) Framework: Bridging the Science-Practice Chasm to Enhance Robust Student Learning},
year = {2011},
bdsk-file-1 = {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}

}

Abstract

Despite the accumulation of substantial cognitive science research relevant for education, there remains confusion and controversy in the application of research to educational practice. In support of a more systematic approach, we describe the Knowledge-Learning-Instruction (KLI) framework. Expressed at a just-right medium grain size, KLI promotes the emergence of instructional principles of high potential for generality, while explicitly identifying constraints and opportunities of detailed analysis of the knowledge students may acquire in courses. Drawing on research across domains of science, math, and language learning, we illustrate the analyses of knowledge, learning, and instructional events that the KLI framework affords. We present a set of three coordinated taxonomies of knowledge, learning, and instruction. For example, we identify three broad classes of learning events: a) memory and fluency processes, b) induction and refinement processes, c) understanding and sense-making processes, and we show how these can lead to different knowledge changes and constraints on optimal instructional choices.

p. 1

One-liner: Defines a theoretical framework for knowledge, learning, and instruction including taxonomies for instructional, learning, and assessment events that occur during the processes of teaching and learning.

Knowledge components (KCs)

Basic unit of learning is knowledge component, e.g.

  • production rule
  • schema
  • misconception
  • facet
  • concept, principle, fact, or skill

unit task level of Newells time scales of human action

p. 14

Instruction makes KC application easier over time

KCs are hierarchical

p. 15

Categorizing KCs by variability

Categorization by application condition and response condition; approximates complexity

  • Constant-constant: facts; studied in research that emphasizes memory p. 18
  • Variable-constant: “conceptual and perceptual category learning”, “artificial grammar rule learning” – rules with many-to-one mappings p. 19
  • Variable-variable: “more complex rule or schema structure” p. 19

Verbal vs non-verbal

  • Whether the application of KC can be readily described in words p. 20
  • Similar to declarative vs. procedural knowledge p. 20
  • Example: Aleven & Koedinger, 2002 describes non-verbal knowledge in geometry p. 21
  • Application: whether to prompt for self-explanation p. 22

With and without rationale

  • Not all-or-nothing p. 23
  • Rationale can be used to “reconstruct a partially forgotten KC, adapt it to unfamiliar situations, or even construct a KC from scratch” p. 23
  • Application: whether to use collaborative argumentation or discovery learning p. 23

See tables for more info and examples

Integrative knowledge component

  • Only applied when integrated with other KCs

    p. 24

Knowledge component complexity

  • Simple heuristic: length of the description of the KC (in English or a cognitive modeling language)
  • Another heuristic: difficulty that students have in applying the KC

p. 25

Other issues

  • Studets may overspecialize conditions or incorrectly generalize conditions p. 28

Assessment events (AEs)

  • Multiple AEs needed to determine that a student has learned a KC p. 19
  • Delayed assessment better than immediate p. 19
  • Fluency and time-limited contexts also a different matter p. 19
  • “Variety in tasks contexts” needed for variable condition KCs (e.g. Aleven & Koedinger, 2002) p. 20
  • Integrative KCs require AEs of supporting KCs with and without the integrative KC p. 28

Examples where different types of KCs have different learning results

  • Rohrer & Taylor, 2006 “the size of the spacing effect declined sharply as conceptual difficulty of the task increased from low (e.g. rotary pursuit) to average (e.g. word list recall) to high (e.g. puzzle)” p. 12
  • wulf2002principles low challenge skills should have an increased load; high challenge skills should have a decreased load p. 12
  • Self-explanations can be effective but can interfere with perceptual task learning schooler1997loss or not make up for instruction time costs wylie2009self p. 21

These may be better understood by applying the KC taxonomy

Instructional event choices

“…instructional principles should refer to KCs rather than to domains”, e.g. use pure practice in math learning when learning algebra “grammar” rules p. 29

More on insturctional principles:

Accelerating learning:

  • Learning how to learn
  • Acquiring deep concepts or foundational skills
  • Increasing cognitive head room, e.g. Hausmann & VanLehn, 2007

p. 32

Fluency gains:

p. 32

Induction and refinement:

Understanding and sense-making:

p. 35

Asymmetry hypothesis: all three learning processes are useful for complex material; memory and fluency more important for simple material. In other words, failures to learn are sometimes failures in memory (Frank & Gibson, 2011). p. 37

Applying KLI theory

Cognitive load vs. knowledge gap theory for worked examples. Backward fading more effective (renkl2002from) or maybe not (renkl2004how) – actually it depends on the type of KC. p. 46

“Expertise reversal effect”: Straight problem solving may be better for more experienced students (Kalyuga et al., 2003). KLI: Beginning students are doing KC induction and sense-making; advanced students, refinement and fluency building. p. 49

Images

[[skimx://koedinger2011knowledge#66|p. 66]]

[[skimx://koedinger2011knowledge#67|p. 67]]

[[skimx://koedinger2011knowledge#68|p. 68]]

[[skimx://koedinger2011knowledge#69|p. 69]]

[[skimx://koedinger2011knowledge#70|p. 70]]

[[skimx://koedinger2011knowledge#71|p. 71]]

[[skimx://koedinger2011knowledge#73|p. 73]]

[[skimx://koedinger2011knowledge#74|p. 74]]

[[skimx://koedinger2011knowledge#75|p. 75]]

ref/koedinger2011knowledge.txt · Last modified: 2014/07/05 00:29 by ryan