Professor Bieber - CIS677 - Decision Process Notes - V2
Simon's IDC Model of Decision Making
- Intelligence: identify problems; collect and analyze information about the problem
- Design: design several possible scenarios/solutions
- Choice: choose the solution
Decision Process Life Cycle
- search (directed/undirected search of your domain)
- analysis (historical analyses / analyzing formal studies)
- problem formulation
- design (potential solutions)
- choice
- ratification (getting consensus of those affected or in charge)
- implementation (execution, control, coordination)
- documentation (flowchart, decision tree, pseudo code {if then's...})
- feedback (evaluating consequences, determining quality of decision and the process)
Intelligence
Problem Finding
finding the difference between an existing situation and a desired state
e.g., compare expectations against reality using
- historical models - expectations based on past trends
- planning models - expectations based on plans
- organizational personnel models - expectations based on organizational structure
- extra- organizational models - expectations derived from competition, customers & industry
Problem Formulation (ensure you solve the correct problem)
Formulating a manageable problem:
- determine the boundaries of what's included in the problem
- examine changes that may have caused/aggravated the problem
- factor the problem into subproblems
- focus on the controllable elements
Support Needed
Databases:
- societal: economic, social & legal environment
- competitive: marketplace
- internal: capabilities, strengths, weaknesses, constraints of the organization
Need to search these databases for opportunities and problems:
Search Types:
- structured, continuous search (things to examine regularly; periodic reports)
- structured, ad hoc (models & routines available; triggered by exceptions; run when needed)
- unstructured (no set guidelines available; user must be able to make queries and find things through trial & error, changing parameters, etc.) needs flexible system
Standard Reports:
- Summarization
- Comparison (with plans; with competition)
- Prediction/Forecast (on plans; on current situation)
- confirmation (audits)
Design
Designing alternative solutions/responses to the problem.
• Tools to create different solutions:
- optimization under certainty
- payoff matrixes (statistical decision theory)
- utility and indifference curves
- decision trees
- game theory (esp. for negotiating) If I choose this & he chooses that...
- ranking/weighting/elimination by aspects
--- multi-criteria decision analysis (implements multiple aspects)
• Outcome Characteristics
- result with certainty (optimization)
- result with known risks & probabilities (payoff/utility/decision trees)
- uncertainty (probabilities not known) - decision trees, others
Quality Decision Making
DM should to the best of his/her ability & within his/her info processing capabilities:
(1) thoroughly check a wide range of alternatives
(2) gathers full range of goals and implications of choices
(3) weighs costs and risks of both positive and negative consequences
(4) Intensively search for new information for evaluating alternatives
(5) take all new information into account, even when it doesn't support initial course of action
(6) reexamine positive and negative consequences of all alternatives, including initially rejected ones
(7) make detailed provisions for implementation, including contingency plans for known risks
Can a DSS help with all these steps?