Generalisable aspects of expertise

In introduction to Cambridge Handbook of Expertise and Expert Performance, Anders Ericsson reviews several conceptualisations of generalisable aspects of expertise.

He also elaborates some definitions:

Expertise comprises a set of charactersitics/skills/knowledge that distinguish experts from novices and less experienced peers.

Expert performance denotes types of superior reproducible performance of representative tasks of a domain.

In some domains there are no objective measures of these two notions; subjective measures are then used.  These subjective criteria include:

  • recogntion by peers as a reliable source of knowledge/skill
  • authority and status accorded by public or peers
  • prolonged/intense  experience through practice and education

The subjective criteria are often problematic, for example experience, which could mean that difference from novices are a function of repetition rather than superior skill.

Ericsson outlines key issues in expertise development that are currently not well understood and require further research:

  • How experts organise their knowledge and performance?
  • How can efficiency of learning be improved to reach higher levels of expert performance?
  • Why indviduals improve their performance at different rates and why different people reach different levels of final achievement?
  • What are the mediating mechanisms of expertise development?

Source: Ericsson, A. (2006). An introduction to Cambridge handbook of expertise and expert performance: It’s developemnt, organisation and content.  In Ericsson, K.A., Charness, N., Feltovich, P., & Hoffman, R. (Eds.), The Cambridge handbook of expertise and expert performance (pp.3-19). Cambridge, MA: Cambridge University Press.

Developmental level of self-regulatory skills in the workplace

Barry Zimmerman in his article Attaining self-regulation: A social congintive perspective argues that most skills (congnitive and motoric) are initially acquired by observing, reading, or hearing about the performance of skilled social models (teachers, experts, experienced peers, etc). He argues that the socially-conveyed skills become self-regulated through a series of levels. These developmental levels of regulatory skills are:

Level 1- Observation-Vicarious induction of a skill from a proficient model

Level 2 – Emulation-Imitative performance of the general pattern/style of a model’s skill with social assistance

Level 3-Self-control-Independent display of the model’s skill under sturctured conditions

Level 4 -Self-regulation-Adaptive use of skill across changing personal and environmental conditions

Zimmerman says that there is evidence that the speed and quality of the development of self-regulatory skills “can be enhanced significantly if learners proceed according to a multilevel developmental hierarchy” (p. 31). He then describes an unpublished study by Kistantas, Zimmerman and Cleary* who compared the development of dart skill by novices who learned initially from modelling (a skilled dart player demonstrated dart throwing strategies and provided feedback on a selective basis) with that of learners who initally learned from enactment.  The study found that learners who had the benefit of modelling “significantly surpassed the dart skill of those who attempted to learn from performance outcomes only” (p. 31). And “learners who received feedback learned better than those who practiced on their own, but the feedback was insufficient to make up for the absence of vicarious experience” (p.31). Learners exposed to strategic modelling “showed higher levels of self-motivation according to an array of measures such as self-efficacy and intrinsic interest than students who realied on discovery and social feedback” (p.32).

It would be interesting to conduct a similar study in the context of self-regulated learning in the workplace (non-instructional, non-formal learning), in addition to extending it to cognitive rather than only motoric skills.  It would also be interesting to study to what extent exerienced peers can facilitate development of self-regulatory skills in the workplace during levels 1-4.

* Kistantas, A., Zimmerman, B., & Cleary, T. (1999). Observation and imitation phases in the development of motoric self-regulation. Unpublished manuscript. Graduate School of the City University of New York.

Cultural Transmission theory and diffusion of innovation in education

Kai Pata is discussing the applicability of Cultural Transmission (CT) theory to explaining the use of artefacts in web communities. The key idea is that next to genes, culture is another mechanism for heredity and replication.

CT could be a useful framework for understanding the processes underpinning diffusion and adoption of innovations in education (eg innovative teaching practice). It is also interesting to explore the links with memetics and evolution of social systems.

Types of knowledge work: Differences between experts and novices?

Thomas Davenport in his “Thinking for a Living” (Chapter 2) offers a set of classifications of the types of knowledge work.

The first one is a martix structured around complexity (C) of knoweldge work (ranging from routine to interpretation/judgement-based) and level of interdependence (I) requried (from individual cator-based to reliant on collaborative groups). It includes transaction model (low C, low I); integration model (low C, high I); collaboration model (high C, high I), and expert model (high C, low I).

I am wondering if within each of these general types of knowledge work, there are differences in the nature of tasks that novices and experts carry out. Take for example the collaboration model. Davenport claims that knowledge work of this type is characterised by being improvisational, highly reliant on deep expertise across functions, and dependent on fluid deployment in flexible teams. He uses investment banking as a typical example of this type of knowledge work.  But how likley is a novice investment banker to have such deep cross-functional expertise?  Novices will probably start off by doing more routine and process-reliant tasks (the key characteristics of transaction and integration model), which will increase in complexity and level of interaction as their skills and expertise develops. Davenport’s classification model doesn’t seem to reflect a developmental trajectory that exists within each type of work.

We are currently using this classification within a research study in a global multinational company. It would be interesting to see if the upcoming surveys in each of the specialist testbeds within this company will show any differences between novices and experts in relation to the nature and the types of the knowledge work.

***** UPDATE on Oct 2, 2011: This study has now been published in the Journal of Knowledge Management. The full reference is:

Margaryan, A., Milligan, C., & Littlejohn, A. (2011). Validation of Davenport’s Classification Structure of Knowledge-intensive Processes. Journal of Knowledge Management, 15(4), 568-581.