Reference no: EM132352645 , Length: word count:3000
Assignment -
Instructions - Produce a report of a minimum of 3000 words that addresses the following questions:
1. Identify and outline a decision problem for an organisation to choose raw materials from 5 new suppliers. Once the decision problem have been identified. No need for the name of the potential supplier to be stated just refer it as Company A,B,C,D, and E There is an example below that can provide further guidance on the structure of the report.
2. Application of the simple multi-attribute rating technique (SMART) to the outlined decision problem. Each stage of the analysis should be not only clearly defined but also presented and applied. Use tables and graph were necessary. Fictional data can be used when allocating any other costs for the company to provide the service e.g transportation cost of the carton box from the chosen company. Provided that it makes sense. The stages and explanation of each stage is stated below. Do not explain each stage, but rather apply it in the context of the decision problem.
Stage 1: Identify the decision maker or decision makers in the above case study.
Before the formal decision making process can begin, it is important to understand who the involved parties will be. The first stage of the SMART method identifies who the decision makers will be.
Stage 2: Identify the alternative courses of action
Within the second stage of the SMART method, decision makers identify (5) alternative courses of action within the decision problem.
Stage 3: Identify the attributes that are relevant to the decision problem
Following the initial two stages of the SMART method, the decision makers and alternative courses of action have been identified. The next stage of the method is to determine the (5) attributes that the decision maker considers to be the most relevant to the decision problem (Choose any 5 attributes). The value tree to choose the supplier is to be illustrated here. I have attached an example of a value tree prepared for another decision problem.
Stage 4: Measure the performance of alternatives relative to identified attributes
The initial three stages of the SMART method focus towards identifying the possible courses of action and the various attributes under concern within the decision problem. Stage 4 then relates to finding out how well each of the identified courses of action then perform in relation to each of the identified attributes. Through developing a well-structured value tree within stage 3, this process may be simplified through focusing on the lowest level attributes and identifying appropriate variables that represent each attribute. This allows for the development of a quantifiable measure of each attributes in relation to the identified courses of action. If an attribute cannot be easily quantified, then an interval ranking may be applied.
For attributes that cannot be represented by easily quantifiable variables, a direct rating approach may be used. Within the direct rating approach, the possible courses of action are listed and then ranked according to a selected attribute. The highest ranked course of action is given a value of 100; the lowest ranked course of action is given a value of 0. The remaining options (courses of action) are then rated so that the values between options provide an indication of the strength of preference. For the initial set of value it is possible to check for consistency through comparing the values between options. Following this comparison, the values of options may be revised.
For attributes that can be represented by easily quantifiable variables, value functions may be used to provide a measure of performance between alternative options (courses of action). Through this, preferences can be translated on to a 0-100 scale and compared. Dependent on the quantified value of an attribute and the preference of the decision maker, the available options first are ranked.
The best performing option is given a value of 100, and the lowest performing option is given a value of 0. The best performing option is presented in mathematical notation as v(value) = 100. The lowest performing option is represented as v(value) = 0. It is then necessary to find the values of options that fall between the most and least preferred options. This could be achieved through the use of direct rating, however as the attributes are quantified, it is possible to derive their value as a value function. This achieved through the bisection method.
Within the bisection method, the decision maker is required to identify a midpoint between the least and most preferred option values. This is represented as v(value) = 50. Following this quarter-points are identified. The first quarter-points falls between the midpoint and the least preferred option, represented as v(value) = 25. The second quarter-point falls between the midpoint and most preferred option, represented as v(value) = 75. The identified values can then be plotted on a graph and used to estimate the values of options. This is shown in the figure below.
Stage 5: Determine the weighting of each attribute
Once the values have been developed for each of the attributes, these then need to be combined so that the overall effectiveness of a particular course of action can be measured across all of the attributes. This is achieved through assigning a weighting to each attribute. Through this attributes considered more important will carry more influence within the decision analysis.
In the simplest form, this may involve assigning values to reflect the relative importance of the attributes. The difficulty with this approach is that the weighting of attributes may not take into account the range between the least and most preferred options on each attribute. Meaning that the end weighting of an attribute could become skewed. For example, if two potential courses of action or options measure very closely on a particular attribute, then it is unlikely that these options will be considered important within the decision analysis.
To avoid this issue, swing weighs are used within determining the weighting and subsequent influence of an attribute. These are developed through the decision maker comparing the change from the least preferred to the most preferred value on one attribute with a similar change in another attribute. Swing weights can be developed through utilising the developed value tree of the identified attributes. The decision maker then considers a hypothetical situation in which all of the identified attributes are set to their least preferred levels. Selecting one attribute at a time, the decision maker identifies which attribute they consider to be the most important. After this, the decision maker then selects and ranks the remaining attributes in order of perceived importance.
The attribute considered the most important is then given the highest weighting. Through comparing the 'swing' between the remaining attributes, the decision maker may then develop weighting for each of the attributes. Following the weighting of attributes, it is necessary to normalise the results.
Stage 6: For each alternative, take a weighted average of the value assigned to that alternative. Use ROC weights (Roberts and Goodwin, 2002) to determine the weightage of each attributes.
At this stage of the SMART method, the decision analysis has outlined how well each course of action or option has performed on each attribute. Attributes have also been weighted to allow for a comparison between allocated values. The next stage within the decision analysis is to calculate how well each option performs overall.
This is achieved through following the additive model. This method assumes that there is mutual preference independence between attributes. Meaning that choice based on one attribute should not influence the choice based on another attribute. This involves adding the options weighted value scores together to obtain a measure of the overall benefits within a particular course of action. This means that the combined value of particular attributes is greater than the sum of the individual values.
Stage 7: Make a provisional decision ( Choose a Company and give an explanation why the company is chosen)
Following the information provided through the previous stages, the decision maker may develop a provisional decision between the available courses of action or options. Based on the raking and weighting of various attributes, the decision maker can then compare various options based on their perceived importance or influence. Within a decision problem, if a particular option performs better than another option across all of the identified attributes, then it is said to dominate the lower performing option. Options that underperform across the identified attributes are said to be dominated (outperformed). These options may then be removed from consideration, and the remaining options are said to lie on the efficient frontier.
Stage 8: Perform a sensitivity analysis
Sensitivity analysis is used to examine how robust the choice of an alternative is to changes in the value applied within the analysis. The decision maker can then review the impact of changes of a certain attribute on an identified option. The ranking and weighting applied by the decision maker may also be subjective and carry a level of uncertainty. Through conducting a sensitivity analysis, the decision maker is able to identify how sensitive the ranking of identified options are in relation to changes in the identified attributes. This may also provide an insight into which attributes are essential when comparing possible alternatives.
To support the sensitivity analysis, decision makers may draw a graph of the alternatives with respect to the different weights assigned to a particular attribute. This may provide a visual representation of the influence of changes to the value of an attribute.
3. Discuss and detail the strengths and limitations of your analysis in the context of your decision problem
Word length: Minimum 3,000 words excluding references.
Attachment:- Assignment File.rar