DECISIONAL GUIDANCE - Human-Computer Interaction and Management Information Systems: Foundations - page 83

Guidance Provided
Informative guidance included information about the various methods. Suggestive guidance included
conclusions about the datasets and recommendations for which methods to use or not. Four different
classes of guidance were provided by crossing informative and suggestive guidance with predefined
and dynamic guidance.
Guidance provided displays (tables or graphs) that matched the task type (symbolic or spatial).
Suggestive guidance for one task helped “decision makers develop the 'correct' causal map of their
decision environment” and for another helped “decision makers assess relevant dimensions and to
develop the correct SWAT grid.” Much of the suggestive guidance was in the form of feedback that
corrected users. Informative guidance provided the same basic information but did not offer
suggestions concerning the correct answers and did not guide the user to select relevant strategies.
Guidance made assignment suggestions based on a decision rule.
Guidance provided information about “the usefulness of different product features, as well as the
potential cost of having them.” General advice about how to approach a given question was also
provided, but specific suggestions were not.
Guidance provided suggestions, and corrected errors, to facilitate following rules for good database design.
Informative guidance provided scores for each forecasting method on six attributes. Suggestive
guidance allowed users to weight the attributes, calculating an overall score for each method.
Guidance provided feedforward and cognitive feedback to support group cognition and breakpoints in
group interaction.
Guidance of various forms for steering an analyst through a specific systems development methodology.
Guidance for the faithful adoption of heuristics through GDSS features, training, and facilitation,
instantiating and testing adaptive structuration theory (DeSanctis and Poole, 1994).
Informative and suggestive guidance for structuring and for executing executive decision-making
processes with a focus on reducing human information-processing biases.
These two examples—and finding more is easy—suggest that categorizing guidance as
informative or suggestive may be overly simplistic. This dimension of the typology needs a third
category defining a middle ground between informing and suggesting.
Why does it matter if we have two or three categories? For practitioners, the typology repre-
sents a set of design choices, and making the distinction shows more fully the set of options. For
researchers, the primary goal is to understand the behavioral effects of guidance. Guidance that
explicitly recommends might be more likely to be followed than that which more subtly implies.
But the explicit approach might lead to less user learning. These are conjectures, but the distinc-
tion needs to be made so the questions can be studied empirically. Forcing the in-between cases
into one category or the other would likely be problematic when we try to generalize from a col-
lection of studies. For instance, we can conclude from this study that one of these guidance mech-
anisms led to more learning than the other, but given the classification issue, we would not want
to conclude that informative guidance leads to more learning than suggestive guidance.