A conceptual model or paradigm
refers to a specific way of think about something. It usually involves
identifying some particular factors or variables, defines the relationship
among them (i.e., "if such and such happens to X, then this affects Y in
this way..."). While it is tempting to construct simple cause-effect relationships,
human systems are usually much more complex than this. Most events have
multiple factors influencing them, and the feedback loops of information
can create high levels of complexity, ambiguity, and even chaos. Since
real world systems are often unmanageable, mental models are an attempt
to extract the most important components of them and build a simple representation.
The purpose of model building
is to more clearly communicate to an audience/reader what the key components
and relationships are of an event or process. It also enables generating
more questions and testing of hypotheses (i.e., "If X, then Y").
There are at least three
different kinds of models that can be built:
For example, Fiedler's Contingency Theory of Leadership might be expressed
in this way: "The effectiveness of a leaders expression
of a task or relationship oriented style is dependent on the unique combination
of task structure, position power, and leader-member relations present
in a situation." This model identifies the variables of outcome
(effectiveness), leadership style (task or relationship), and contingency
(task structure, position power, leader-member relation). It proposes that
under certain combinations, a specific style is preferred in order to generate
a more effective outcome.
To build a model...
Mathematical models are often refinements of verbal models in which the
variables can be quantified. For example, if leadership performance (LP)
is the desired outcome it might be accurately predicted by a combination
of factors identified by multiple regression. If it was predicted by intelligence
(IQ) scores that had a (beta) weight of .47, a problem solving (PS) weight
of .67, a past experience (EX) weight of .58, and a biographical relevance
(BR) rating of .46, the following regression model could be proposed:
Visual Models. Visual
models can be used to incorporate important dimensions that are difficult
to verbally describe or are too qualitative to be rendered quantitatively.
They are also often symbolic (i.e., the shapes often communicate as much
as what dimensions represent) and can be taken in at a glance. For example,
Peter Senge's model of the five disciplines of leadership (i.e., team kearning,
shared vision, mental models, personal mastery, systems thinking) are combined
in a pyramidal form in which they are refined and integrated as one developmentally
approaches the top of the pyramid.
When working with visual
models consider the following:
Consider the overall process
that you want to model; the big picture.
Identify the feasible portion
on which you want to focus your model
Identify the key components
State the relationships these
components have with each other
State (words, formula, graphic)
the above in a way that most clearly and accurately represents the idea.
More than three dimensions
(e.g., hypercube) is more than most people can follow; keep it to two or
three, or perhaps sequence a string of models
Use the natural implications
of a shape: a pyramid at the top may show integration, a decision tree
increases its branching differentiation, a spiral reflects cycles and recursion,
Use shading, coloring, patterning,
proximity, and arrows to show relationships and qualities.