updated 2007. 3. 5
//    g e n e r a l   t h e o r i e s   w e   g e n e r a l l y   n e e d     // chapter jump
ANDREW
WALENSTEIN

walenste@ieee.org

Research Scientist

Center for Advanced Computer Studies
University of Louisiana @ Lafayette

Chapter 6: CoSTH: a Hierarchy of Support Theories

Theories of cognitive support postulate explanations for why and how artifacts can improve thinking and problem solving. Two ingredients are indispensible for producing any such theory:
  • a description of some facets of cognition that can be improved,
  • a description of ways in which artifacts can improve that facet.
These are precisely the two things that HASTI (chapter 5) and RODS (chapter 4) are intended to provide, respectively. HASTI gives ways of modeling joint cognitive systems and RODS enumerates ways of changing the computational structures of those systems for the better. With these two, one can pump out a number of cognitive support theories of varying types and varying degress of specificity. This chapter extracts a hierarchy of cognitive support theories from these two resources. This hierarchy is named "CoSTH", for "Cognitive Support Theory Hierarchy".

Five interesting results emerge from generating the hierarchy of theories.

  1. First, the collection of theories is itself an interesting result. There are 16 leaf nodes (ie, most specific theories), and 16 other possible refinements are proposed. The theories can be used to classify a wide variety of cognitive support identified in chapter 3. It is a powerful collection.

  2. Second, the hierarchical organization of the theories is a powerful organizer. The hierarchy structures the theories by hierarchically collecting together similar specific theories into more abstract theories. For instance, there are 11 types of distribution (see chapter 4) identified, 2 sub-types, and 8 sub-sub-types. This hierarchical structure helps structure the theories so they are more easily understood, but, significantly, it also allows hierarchical classifications of tools according to the types of support they provide.

  3. Third, the hierarchy is very thin at the top. The topmost level defines just three different types of cognitive support (one of which is distribution, for instance). This means that all of the different varieties of cognitive support are parsimoniously explained using a scant three basic principles.

  4. Fourth, many tools contain a complex composition of different types of cognitive support. In fact, it appears that many of seemingly vast variety of types of supportive artifacts can be explained by decomposing the types of support into simpler "factors". This is somewhat analogous to the situation in chemistry, where the uncountable number of types of molecules can be decomposed into a small, finite number of different atom types. The enormous variety of different types of molecules hides an underlying simplicity. The implication is that complicated tools can be factored or decomposed into "atomic" support types. This paves the way for principled comparisons of tools.

  5. Fifth, not all compositions of support may be maximally beneficial: many special compositions of support appear to be mutually supportive. This might be a direction for future research on what compositions are most synergistic.

The chapter presents each node in the hierarchy, gives a description, and gives one or more examples of each type. For instance "to do" lists are given as an example of a type of distribution. The examples form an illustrative "cookbook" of design ideas. This cookbook is indexed by type of cognitive support being provided rather than by surface features. In addition, the ways in which support types compose is examined. Several types of complex compositions from the literature are decomposed to expose individual facets of support. Comparisons are made to other theoretically-motivated collections of cognitive support theories, and to other ways of categorizing tools.

[ up ]
thesis overview
1
2
3
4
5
6
7
8
9
 
[ up ]thesis overview 1 2 3 4 5 6 7 8 9
chapter jump