Term
|
Definition
-structuring & clarifying complex decisions I.e. When there are Many usually conflicting criteria
Step 1- developing a set of criteria & options
step 2-performance rating of a set of options: shortlisting, ranking & choosing " good " options |
|
|
Term
Substantive & bounded rationality
SR becomes more practical;
-the clearer & less controversial the set of priorities & objectives
-the better defined the set of options or actions
-the better understood or more predictable the relationship btw action & outcomes
-the less extensive the trade-off between competing objectives. |
|
Definition
-Is the rational choice that logic & the Laws of Probability will lead to the best & most preferred set of outcomes
-substantive rationality is the basis of decision analysis chap 6, where the aim is to maximise " subjective expected utility "
But
In practice Decision makers cannot be sure of identifying the best option as may not
be possible / too costly / time consuming / risky or unethical to obtain the necessary information
Or
Even if DM have the info may not have the capacity to process it or preferences may change
Thus most DM work with BOUNDED RATIONALITY
I.e bounded by complexities of the world they live in & limited capacity for analysing it |
|
|
Term
|
Definition
-not only important to seek good outcomes, but also the means by which these are the outcomes achieved I.e.process, also important
-here the test is not whether the best outcome Is achieved but how closely the decision maker follows particular rules or procedures
-these procedures may be;
a) part of the institutional culture, such as rules of thumb that have produced good decision making in The past or
b)based on an appeal to accepted social principals such as transparency, rights & participation.
-For priority setting in Health care "good procedures"are those that are justifiable in terms of 1) how far it is driven by factors that fair minded people agree are relevant,2) it's public acessibility, 3) whether there are mechanisms for appeal, 4) whether adherence to the 1st three items is audited.
-in contrast to subst ratio, in proced ratio calculating or estimating the expected outcome is not an essential step
-SR stresses rational choice whereas PR stresses rational choosing |
|
|
Term
|
Definition
-subst ratio is the Ideal but where it is not possible,decisions makers may satisfice
-This means that instead of evaluating every option, they identify & evaluate successive options until they find the one that meets a set of pre-specified standards: this option chosen bcse good enough despite not shown to bethe best
-this has elements of both substantive & procedural rationality |
|
|
Term
|
Definition
-Multiple criteria decision analysis:
Provide help in the form of how rather than what
-strategic choice can be through the design of visual Aids |
|
|
Term
| Creating a Performance matrix or consequence table |
|
Definition
-is construction of the framework of options & criteria in form of grid/table
-useful as a first step towards identifying the info needed to make a decision
-table organises & presents the info
-activities involved to do this are;
1/ identify the decision context: overall aim & stakeholders( professional groups, patients, neighbour HCP, funders, hosp managers) affected
2 /develop a set of criteria; start with a list of objectives (open-ended/specific) , turn them into criteria -brainstorming&d/w DM
3/identify the decision options thru brainstorm / using problem structuring methods ( etc SODA/SSM, SCA, RA & DT)
4/devise measurement scales & scoring systems for each criteria -scale depend on data available, signif of dec, time & Resouces for data collect.
5/rate the options according to measurement scale. Rating -subjective judgement/direct observation/data analysis/ estimated from a model
-Repeat these steps until satisfactory framework is achieved
-Order of task 2 and 3 maybe reversed |
|
|
Term
| Creating a Performance matrix or consequence table |
|
Definition
Scenario 2: whereby options are already determined for you, generating a criteria is as follows ;
1/list adv & disadvantage option 1 compared to existing situation
2/for each adv & disadv identify a corresponding criterion ( variable not attribute)
3/ list the adv & disadv of option 2 compared to the existing arrangements & option 1
4/Add to the list of criteria if need be
5/ continue with the options until a new criteria arise
6/ finally, shorten list of criteria by removing any w/c essentially measure the same thing or w/c do not descriminate
Notes;
a)any anslysis of the performance matrix is easier if the criteria are ' preference independent' I.e no adv/disadv from a particular combinations
b)difficulty -objectives / criteria maybe abstract & cannot be measured directly, thus
-either options r rated making explicit judgements-impact as good/poor or
-operationalise the abstraction using one/more indicators
-hierarchly / " value tree" e.g cost split one-off or recurring |
|
|
Term
| Analysis of the performance matrix |
|
Definition
-when many criteria & for example options rated as good/ fair/poor
-to shortlist options there are 3 main categories of approaches:
1st approach Satisficing: setting standards for what is good enough for each criterion, number characteristics that every option must have
2nd approach weighing:
-first transform into scores, so the highest score means more preferred for each criteria,
-then multiply each score by the relevant weight &
-add all the weighed scores for each option together so as to give an overall score-called LINEAR ADDITIVE WEIGHTING
1/Weighing adv; consistency in decision making
2/Disad: difficult to use as
A) scores have to have equal interval strength of preference scales
B) must take into account ( in scores or criteria if importance ) if difference criteria are rated in difference scales
C) preference rating on each criteria should be mutually independents
Notes
DM may disagree with regards to weights or choose those that produce the answers that they want
3rd approach sequential elimination:
-option A is said to dominate B if scores better than, or as well as option B on all criteria, eliminate option B w/o makjng any value judgement
-transparent & undemanding ; only requires an agreement about the broad ranking of each option on each criteria, not on the relative option of each criteria
-useful step +possible to eliminate more options by 1/examining specific pairs & debating possible trade-offs - PRACTICAL dominance
-limitations :
a)results v sensitive to performance rating & thus to diff in judgement- based scores or error in the data
b)a shortlist will include the best options but not necessarily consist of the best 5 options
c)generally there are few dominated options unless large number of tied scores -e.g if scoring system insensitive
4 th more approaches & ways of going further :
-Evan swaps
-Outranking
-Analytical hierarchy process-importance weights based in pairwise preferences |
|
|
Term
| Pros & cons of Multiple decision analysis |
|
Definition
1-Advantages
-it's open & explicit
-the choice of objectives & criteria made is open to analysis and change if they are felt to be inappropriate
-scores & weights when used are explicit & developed according to established technique
. Can be cross referenced to other sources if info on relative values & amended if necessary
-performance measurements can be sub-contracted to experts instead of leaving it in hands if decision makers themself
-can provide an importance means of communication w/in decision making body & later btw that body & wider community
-scores & weights are used w/c provides an audit trail
2-disadvantages :
Difficult to arrive at an agreed & properly scaled set of preference scores & arriving at an agreed set of criteria weights
Note: sequential elim or pairwise comparison is
-easier to do
-allows good progress towards short listing w/o having to make trade offs
But
-Later stages of elimination process controversial
-Sensitivity analysis more diff than weighing methods |
|
|
