Decision Theory Flashcards
Decision
A decision is the choice between at least two alternatives, one of which can be the so-called injunction alternative.
Constitutive Decisions
Constitutive decisions are fundamental and far-reaching decisions that
– Are difficult to revise and
– define the long-term framework for all subsequent operational decisions.
- Location
- Legal Form
- Intercompany relationships
Decision - making field
Action space (actions a1) Action (a) (alternative action): independent approach that can be chosen by the decision maker to achieve the objectives. ---> Result function f (Results e1 = f(a, z) result (e): unrated result of the implementation of an action (a) at certain environmental conditions (z). Also state space Environmental conditions zj ------> results matrix e (a,z) Result per action and environmental status
Inter dependence relations
• Neutral goal relationship
– The fulfilment of goal k1 has no influence on the
fulfilment of goal k2.
– Example: Increasing job satisfaction and reducing
energy consumption in production
• Complementary goal relationship
– The action aimed at achieving goal k1 shall at the
same time support the achievement of goal k2
– Example: Improving company social benefits and
increasing job satisfaction
– Extreme case: Complete congruence of goals: goals
are identical
• Conflicting goal relationship
– A better fulfilment of goal k1 worsens the
achievement of target k2.
– In extreme cases, goals are mutually exclusive:
target antinomy
– Example: Goal triangle
Goal Triangle
Result : Internal - quality; External - Customer value;
Time: Internal - Throughput time; External: Delivery time;
Effort: Internal - Cost; External - Price;
Operationalizatin of goals
- Content: Qualitative, Quantitative
- Magnitude: extremation fixation • minimum • maximum • goal corridor
- Time reference: point in time, period
- Scope
Decision model
Certainty: decisions with only one single, known environmental state (target weighting, Lexicographical)
Uncertainty: only known, that any of the possible environmental states will occur, not probabilities of occurrence are given (Maxmin rule, Hurwicz rule, Savage Niehans rule, Laplace Rule);
Risk: several environment states, probabilities of occurrence of different environmental states, that can be determined. (μ principle, (μ,σ)-Principle)
Lexicographical order
• Goals are prioritized according to the preference of the decision maker
• Can be a useful method for dominant, critical goals
• Procedure:
– Choose the action that best meets the most important goal
– If there is no clear solution, choose the action that best meets the second most
important goal.
– If again there is no clear solution, choose the action that best meets the next
most important goal.
– … Continue until a clear solution is found.
Minimum-Regret-/Savage-Niehans-Rule
Procedure:
– Determine benefit loss when choosing an action
– The assessment factor is the maximum loss of benefit
– Decision rule: “Choose the action with the minimum appraisal size”.
Laplace Rule
Assumption: All environmental conditions have the same probability of occurrence.
• Decision based on average utility values
Procedure:
– The assessment value results from the average of the utility values.
– Decision rule: “Choose the action with the maximum appraisal size”.
μ/Bayes Principle
Assumption: Risk neutrality of the decision maker.
Procedure:
– Utility values of the various environmental states of an action are weighted with
the probability of occurrence (“Weighted average of environmental states”).
– The assessment variable results from the respective summation
– Decision rule: “Choose the action with the maximum appraisal size”.
(μ, σ)-Principle
Assumption: higher expected values are associated with higher risks
Fields of decision theories
• Descriptive Decision Theory
– How decisions are made in reality
– Objective: To forecast how decision-makers will behave in the future and how
this behaviour can be influenced.
• Prescriptive Decision Theory
– Guidelines on how to proceed in certain decision-making situations