DECISIONAL GUIDANCE - Human-Computer Interaction and Management Information Systems: Foundations - page 79

On-Demand or
Automatic Principal Findings

• Guidance outperformed no guidance for decision quality, user satisfaction, learning,
and decision time (not total time)
• Suggestive guidance outperformed informative guidance for decision quality and
decision time
• Informative guidance outperformed suggestive guidance for learning
• Dynamic guidance outperformed predefined guidance for decision quality, learning,
and decision time

On-demand

• Guidance outperformed no guidance
• Most subjects preferred to make their own choices vs. using “good pre-selected
displays”
• Most subjects preferred to have the software advise them vs. choosing their own
displays without advice
• Effects on response time depended on task type

Automatic

• Matched displays outperformed mismatched displays
• With guidance, field dependent individuals and field independents performed
equally well

Automatic

• Less complex tasks: suggestive guidance outperformed informative guidance
outperformed no-DSS
• More complex tasks: informative guidance outperformed suggestive guidance and
no-DSS, which tied

Automatic

• Guidance outperformed cognitive feedback (CF) and no guidance/feedback
• Users of guidance and CF equally satisfied, more satisfied than users without
guidance and CF

Automatic

• Guidance increased trust in integrity On-demand

• Guidance outperformed no guidance/no restrictiveness
• Guidance and restrictiveness performed equally well
• Restrictive system perceived easier to use than one with guidance

Automatic

• Users employed different choice strategies when given different forms of guidance
(suggestive vs. informative)

Automatic

• Groups with guidance had greater model understanding and greater decision time
• Groups with guidance had better perceptions of the group decision process and
outcomes, except for confidence
• Groups with guidance had better perceptions of the MCDM GDSS

Automatic

Performance

Eight of the studies addressed individual support, while the remaining one focused on group sup-
port. Six of the eight studies of individual support used some measure of outcome quality. Four of
the six (Wilson and Zigurs, 1999; Parikh et al., 2001; Huguenard and Ballou, 2001; Antony et al.,
2004) reported comparing various kinds of decisional guidance with computer-based systems
containing no guidance. All four found significant, positive contributions of decisional guidance
to performance quality.