Structural Complexity and Effective Informing

Introduction The idealized informing system consists of a sender supplying information to a client through a channel in support of a task (Cohen, 1999). For such informing to be effective, the information being supplied must impact the mental model of the client. We refer to this model as a problem space (Gill & Hicks, 2006). Given this general framework, it makes sense to speculate that a qualitative understanding of the sender's and client's respective problem spaces could lead to insights regarding both the nature of the information that will be most effective in informing the client and its presentation. Such insights, in turn, could then be translated into rules that could be applied towards achieving more effective informing--an objective that is at the very core of the informing sciences. It has been proposed that task complexity may be an important contributor to the informing process (Gill & Hicks, 2006). Of particular interest is structural complexity, which defines task complexity in terms of the nature of the task specific knowledge in the problem space. As such, it changes as the performer learns more about the task. This learning aspect of structural complexity makes it a good candidate for giving us insights into the process of informing. The present paper adapts a structural complexity model, first proposed for use in matching task characteristics to expert system tools (Gill, 1991), to the general informing problem. It then analyzes how different levels of problem space structural complexity virtually demand different qualitative types of informing. We begin by presenting some definitions and then introducing a model of structural complexity, reviewing relevant literature as the model is developed. We next consider what types of informing are most likely to succeed at each level of structural complexity. Finally, recommendations deriving from the hypothesized relationship between informing and structural complexity are presented. Definitions Prior to developing the structural complexity model to be used in this paper, it is useful to present a number of definitions. We begin with how we will be using the term task: Definition: a task is a set of assigned a) goals to be achieved, b) instructions to be performed, or c) a mix of the two. (Hackman, 1969) We next define the term problem space as follows: Definition: A problem space is a representation of the cognitive system that will be used to perform a task "described in terms of (1) a set of states of knowledge, (2) operators for changing one state into another, (3) constraints on applying operators, and (4) control knowledge for deciding what knowledge to apply next." (Card, Moran, & Newell, 1983, p. 87) To organize our problem space, we will describe the set of knowledge states as the state space and the set of operators, constraints, and control knowledge as the operator space. Because tasks without goals--item (b) in our task definition--tend to be highly structured (almost as a matter of definition), we will also assume the existence of a goal space, containing one or more operators that allow the task performer to assess the fitness of the state space with respect to task goals and to determine relative progress towards these goals. Structural task complexity, or structural complexity, refers to the degree to which a task is performed using task specific (as opposed to general purpose) knowledge, operators, and goals. Under a structural complexity definition, low structure is more complex than high structure, i.e., unfamiliar tasks are more structurally complex than routine tasks. To avoid confusion with the many conflicting definitions of task complexity (see the complete review in Gill & Hicks, 2006), it should be noted that structural complexity differs from other widely used task complexity definitions in a number of important respects, including: * Unlike objective task complexity (Wood, 1986)--which is defined to be a function of the number of task components, their degree of inter-relationships, and the degree to which these relationships are dynamic--structural complexity is not a property of the task itself but of the problem space used to perform the task. …