Mercurial Hosting > go2
view src/RDMCh5.html @ 0:0e911cd3fd2a
add content
| author | Franklin Schmidt <fschmidt@gmail.com> |
|---|---|
| date | Sun, 02 Apr 2023 10:30:03 -0600 |
| parents | |
| children |
line wrap: on
line source
<html> <!-- Mirrored from users.eniinternet.com/bradleym/RDMCh5.html by HTTrack Website Copier/3.x [XR&CO'2014], Sun, 06 Nov 2022 06:59:35 GMT --> <head> <title> Reasoning And Decision Making - Milton N. Bradley</title> <link rel="stylesheet" href="style2.css"> </head> <body bgcolor="#E0FFFF"> <a name="top"></a> <center><br><br> <font class="booktitle"> Reasoning And Decision Making</font> </font> <br><br> <h1> <font class="chaptitle"> © Milton N. Bradley 2010</font> <hr></h1></center> <br><br> <center><a name="Chapter5"></a> <font class="chaptitle"> <strong>Chapter 5 - Developing The Problem Solution</strong> </font><br> </center> <br><br><br> <font size= +1> <strong>Part 1 - Assuring Inclusion of All Relevant Parameters</strong><br><br> When we’ve finally decided that it makes economic and otherwise practical sense to solve our problem, the next question to be decided is how to systematically and thoroughly go about actually developing its solution consistent with the principles identified earlier. To that end a number of related techniques have been developed for quickly and accurately focusing on those system aspects, parameters and constraints that are most likely to be important, and to which attention is most likely to pay off. <br><br> <ul> <li> <strong>Ways To Attack A Problem</strong><br><br> <ul> <li><strong>Entry Points</strong> <ul> <li><strong>The Beginning.</strong> The conventional approach. <li><strong>The End.</strong> Approximate/guesstimate an answer, then work backward to try to figure out how it could be obtained. <li><strong>Somewhere in the middle.</strong><br><br> </ul> <li><strong>Rival Hypotheses.</strong> Conjure up different theories about the nature of the problem and how to solve it, and see where each leads.<br><br> <ul> <li><strong>Modeling <ul> <li>Modeling Objectives: <ul> <li>Make an idea concrete. <li>Reveal possible relationships <li>Simplify the complex. <li>Clarify the problem. </ul> <li>Model Categories <ul> <li>Conceptual. <li>Structural. </ul> <li>Model Types. <ul> <li>Visual. <li>Physical. <li>Mathematical. <li>Metaphoric, Symbolic, or Analogic. </ul> <li>Model Paradigms <ul> <li>System <li>Design <li>Construction <li>Recipe </ul> <li>Specific Model Metaphors: <ul> <li>Garden <li>Machine <li>Symphony Orchestra <li>Human Body <li>Vehicle<br><br> </ul> </ul> </ul> </ul> <li>Conceptual Solution Generating Techniques</strong><br><br> We begin by discussing a generic idea generating technique, which can be applied together with any of the specific technically oriented techniques that follow it.<br><br> <li><strong>BRAINSTORMING.</strong><br><br> In this innovative individual or group process, the objective is to deliberately generate as many unusual solution ideas as possible in a non-threatening, unstructured manner, and then to push each of those ideas as far as possible. To break down preconceptions about the limits of the problem and assure the free flow of the participants’ imagination, during this process no criticism of the proposed ideas is permitted. Only later, when the results of the brainstorming session are analyzed, is it then permissible to evaluate the results and prune the proposed solutions for further brainstorming or the application of other techniques.<br><br> One approach to brainstorming is to “seed' the session with one or more words pulled randomly from a dictionary, as starting points in the process of generating ideas.<br><br> <ul> <li><strong>Group Brainstorming</strong><br><br> Each session should be: <ul> <li>Announced as lasting a fixed (and reasonable) length of time. <li>Moderated and controlled by a designated leader who should: <ul> <li>Try to select participants from as wide a range of disciplines with as broad a range of experience and expertise as possible, to maximize the number and kind of creative ideas developed. <li>Initially define the problem to be solved <li>Stipulate the solution constraints and other criteria to be met <li>Encourage: <ul> <li>an enthusiastic, uncritical attitude among the brainstorming team members so that they will feel free to produce as many ideas as possible, from the solidly practical to the wildly impractical. <li>participants to not only develop their own ideas but also to follow up on those proposed by other team members. <li>succinct expression of ideas, with no elaboration required as to how the proposal might (or might not) work. <li>Encourage participation by all members of the team </ul> <li>Keep the group “on topic” <li>Try to (very gently) steer the group towards the development of practical solutions, while <ul> <li>Ensuring that no train of thought is followed for too long. <li>Visually documenting the ideas developed on a blackboard, easel or similar device, so that all session participants can be “on the same page” at all times. <li>Maintaining a “permanent” written, tape or video record of the session, for later study and evaluation.. <li>Refraining from any sort of criticism which would stifle creativity and cripple the free interchange of ideas essential to any brainstorming session’s success. </ul></ul></ul><br><br> <li><strong>Individual Brainstorming</strong><br><br> Individuals are always free to explore ideas on their own without any fear of criticism or possible domination by other group members. Although individual sessions have the potential to produce a satisfactorily wide range of ideas, they tend not to develop them as effectively as group sessions, probably because most individuals almost necessarily have a smaller range of experience and expertise than even an “ordinary”group can readily muster.<br><br> The result is that group brainstorming is usually superior, because when any individual’s limitations in the development of an idea are reached another’s creativity and experience may be available to overcome that barrier. On the other hand, group brainstorming can be inhibiting if quiet but creative people are allowed to be suppressed or overwhelmed by more aggressive ones, because of ineffective session leadership.<br><br> <li><strong>Mixed Individual and Group Brainstorming</strong><br><br> . Sometimes “the best of all possible worlds” can be achieved by a creative mix of both types of session. This usually takes the form of allowing team members to individually generate a wide (but possibly shallow) range of solutions, and then developing and enhancing these ideas via group brainstorming. The main advantage of this approach is that maximum input can be obtained from creative but shy and/or insecure participants.<br><br></ul> <li><strong>The “SIX THINKING HATS” of Edward de Bono.</strong><br><br> <strong>This is a powerful technique for considering potential decisions from a number of quite different perspectives</strong>, forcing you to transcend the limitations of your normal thinking style to obtain a more rounded view of the problem situation. Rather than simply considering the situation rationally, it also requires you to view it emotionally, intuitively, creatively and even negatively. By thus “touching all the bases”, the probability that any important aspect of the situation will either be overlooked or unnecessarily discounted will be greatly reduced, and the quality of the proposed solution accordingly enhanced.<br><br> <strong>The “Six Thinking Hats” technique is most often applied in group settings</strong>, where it has the benefit of blocking the confrontations that frequently occur when people with different thinking styles discuss a problem. But it is also useful when applied in solo mode.<br><br> <br> <strong>The thinking style for each type of (imaginary) hat is as follows:</strong><br><br> <ul> <li><strong>White Hat:<br><br> This is the rational approach</strong>, with focus on the available data and its analysis. Consider any gaps in that data as well as what you can learn from it, such as trends, anomalies, significant differences, etc.<br><br> <li><strong>Red Hat:<br><br> This is the emotional, intuitive approach.</strong> Consider your own “gut reactions” as well as those of any other participants, especially the ones who may not fully understand or agree with your reasoning or position.<br><br> <li><strong>Black Hat:<br><br> This is the negative approach.</strong> Review the proposed decision cautiously and defensively, considering all the reasons you can think of that it might not work. By thus finding the weak points in an approach you gain the opportunity to eliminate or alter them, or to prepare contingency plans to counter those weaknesses in advance of taking what otherwise might be a costly decision..<br><br> This perspective is one of the real benefits of de Bono's technique, because it avoids getting “blind sided” as a result of undue optimism which prevents you from seeing problems in advance.<br><br> <li><strong>Yellow Hat:<br><br> This is the optimistic viewpoint</strong> that helps you to see all the benefits of the decision and the value in pursuing it, even when everything looks gloomy and difficult.<br><br> <li><strong>Green Hat:<br><br> This is the creative approach</strong>, similar to brainstorming, in which there is little criticism of ideas.<br><br> <li><strong>Blue Hat:<br><br> This is the executive/managerial approach</strong>, switching thinking back and forth into the other perspectives as needed to generate the problem solution..<br><br> <li><strong>A variant of the “Six Hats” technique is to look at the problem from the point of view of different professions (e.g. doctors, engineers, sales directors, etc.), or the specific different participants (individuals or groups) involved.</strong> As is well known, this is certain to produce contrasting or even diametrically opposing viewpoints. The output from the “Six Hats” technique or its variants can also be quantified and then analyzed in precisely the same manner as we show in the techniques described next.<br><br> </ul</ul></ul></ul> <li><strong>Problem Solution Search Techniques</strong><br><br> <ul> <li><strong>General</strong><br><br> These techniques are applicable to any type of problem, and therefore can be used in advance of, in conjunction with, or independent of heuristic and technological techniques. <br><br> <li><strong>Public Solution.</strong><br><br> Publicly posting a request for help on a problem to coworkers, associates, friends, teachers, etc. can get several minds with fresh perspectives working on it independently, in discussion with others, and/or in conjunction with you.<br><br> <li><strong>“Phillips 66".</strong><br><br> This is a form of the Public Solution technique which is useful for simultaneously engaging the minds of a large group in a problem. In its original form, the large group was divided into smaller groups of six people each, and each of those subgroups discussed the problem for six minutes, followed by formulating a single question or proposed solution by agreement.<br><br> There were several inadequacies in this approach, but the main one was the unreasonably short time allowed for discussion. (In six minutes not much can happen because just settling the group down and focusing take some time, and discussing and agreeing typically require much more.) So the revised procedure is now as follows: <br><br> <ul> <li>Each subgroup should consist of 3 to 8 people, with an ideal size of about 5. <li>The assigned task must be made clear and explicit, requesting a decision or conclusion. (Otherwise the output will likely be just feelings, opinions, or questions.) <li>Each subgroup should be allowed at least 20 to 30 minutes to work on the problem. </ul><br><br> <li><strong>The Method of Contradiction.</strong><br><br> Pretend to assume the opposite of what you want to prove and then look for facts that contradict that assumption. By disproving all of the alternative explanations you automatically highlight (and “justify”) the sought after solution, but (unfortunately) you don’t necessarily prove it. (That would only be possible in the relatively rare case in which the two competing explanations were true diametrical opposites.)<br><br> <li><strong>The Method of Outrageous Provocation.</strong><br><br> Begin with a statement that is known to be absolutely incorrect, and then use it as a bridge to a new solution approach. (This is a variant of the Method of Contradiction.)<br><br> <li><strong>The Random Word Technique.</strong><br><br> Select a random word from the dictionary and juxtapose it with the problem statement, then brainstorm about possible relationships.<br><br> <li><strong>Incubation.</strong><br><br> If you’re not making progress toward a solution at an acceptable rate, put the problem aside for a while (overnight, a day or two, several weeks or months, or...) and allow your subconscious mind to work on it. This doesn’t always succeed, but it is sometimes the only path to the solution.<br><br> <li><strong>Case Analysis.</strong><br><br> Research similar cases to see what insights their solutions can offer.<br><br> <li><strong>Heuristic Search Methods</strong><br><br> A heuristic is an empirical, learn-as-you-go procedure which can’t guarantee finding a solution, but which will increase the probability of finding a solution over what would be possible using a purely random or ad hoc approach.<br><br> <ul> <li><strong>Trial and error.</strong><br><br> <strong>Different possible configurations are generated, and after a test of their "fitness" the good ones are retained and the bad ones eliminated.</strong><br><br> The search for a solution proceeds without any sense of choice or calculation of the likelihood that any path will be superior to another. This technique can be made as efficient as possible by being systematic and keeping a record of attempts and their outcomes, so that no path or solution is tried more than once.<br><br> <li><strong>Proximity.</strong><br><br> <ul> <li><strong>Hill climbing.</strong><br><br> In this strategy, the searcher looks one step ahead to see what next action will best move him toward the goal.<br><br> <li><strong>Means-ends analysis.</strong><br><br> This is a set of procedures designed to reduce the distance from the problem state to the goal state through a sequence of carefully identified steps, taking into consideration all of the many different dimensions of difference between those states. It involves creating subgoals (the ends) to eliminate the differences between the current state and the condition for applying a desired operator (the means).<br><br> <li><strong>Fractionation.</strong><br><br> These methods involve breaking the problem into parts and solving each part separately.<br><br> <li><strong>Mixed Scanning.</strong><br><br> This strategy involves alternating between a general overview of the situation and a close-up, detailed, precise examination of a subset of it. The first gives the solver a quick sense of the overall situation, even though that may be somewhat vague and general. Then, given that context, it allows attention to individual parts in exact detail, permitting a close knowledge of pertinent areas to provide information about both potential or actual trouble spots, as well as a better basis for generalizing about the whole.<br><br> <li><strong>The split-half method.</strong><br><br> This method involves starting at the halfway point in a more or less linear system, and checking to see if the problem (or its symptoms) already appears there. If it does, the problem is in the first half of the system. If it doesn't, the problem lies in the second half. Next, the investigator goes to the half of the system where the problem is now known to occur and checks at its halfway point to see if the problem or symptom appears there. Each such iteration successively eliminates another half of what remains of the system as the source of the problem, and the process continues until the problem is located. This is much faster than random checking or simply starting at one end of the system and then proceeding consistently toward the other end.<br><br> <li><strong>Pattern Search</strong><br><br> In selected situations, the presence of patterns may either suggest solutions themselves or remind you of prior cases in which useable solutions were found.<br><br> </ul> <li><strong>Solution Selection Procedure:</strong><br><br> <ul> <li><strong>Develop possible solution alternatives.</strong><br><br> Using as many of the techniques discussed below as are appropriate to the problem being worked on, create solution alternatives over the entire range of options, without concern for feasibility. Enough time should be spent on this activity to ensure that all plausible solutions are uncovered, including non-standard and creative ones.<br><br> Relevant questions to be asked include: <ul> <li>What constraints exist on the solution? <li>What are the available resources for implementation? Are they adequate? <li>Who is the decision-maker, and what are his/her sources of power in the situation? <li>What are possible leverage points for implementation ? <li>Are others involved? If so, are they favorable? If not, is their opposition serious? And if so, can it be changed? </ul><br><br> <li><strong>Evaluate the proposed alternative solutions using the pre-established criteria.</strong><br><br> Techniques like PMI and Cost/Benefit Analysis will come into play here, and alternatives should be directly compared:<br><br> <ul> <li>Eliminate the alternatives which fail to meet essential criteria. <li>Collect Risk/Benefit information for the remaining alternative solutions. <li>Assess the degree of satisfaction each alternative provides. <li>Ensure that the benefits provided are in the areas of high importance. <li>Ensure that high risks and low benefits are not in the areas of high importance. <li>Question if the risks identified are serious enough for an alternative to be eliminated, or whether it’s possible to develop a plan which could remove risk or minimize its effect. </ul><br><br> <li><strong>Select the best scoring/most promising alternative that meets these criteria as the proposed problem solution.</strong><br><br> <li><strong>If no proposed Solution provides sufficient satisfaction, recycle through the entire process again to look for more options!</strong><br><br> <li><strong>When a Solution alternative is finally selected, review and evaluate the potential problems that may arise when this “solution” is implemented, and proceed only if that review is favorable.</strong><br><br> <ul>Relevant questions to be asked include: <li>Does the proposed solution: <ul> <li>Adequately address the critical aspects of the problem? <li>Satisfy the stated objectives? </ul> <li>Is it consistent with organizational and/or other situational realities? <li>Who would be the solution implementation "change agent"? <ul> <li>You? Or someone else? <li>Does he/she have the power, skills, knowledge to be successful? </ul> <li>What is the probability of success? <li>What are the risks if the plan fails? <li>Is resistance to this change likely? <ul> <li>From whom? <li>How strong/serious is this likely to be? <li>Can it be overcome? </ul> <li>Are there any sources of support? <ul> <li>From whom? <li>How important is this likely to be? </ul></ul><br><br></ul> <li><strong>Solution Implementation Procedure:</strong><br><br> <ul> <li><strong>Determine the efficacy and completeness of implementation.</strong><br> Omission or improper completion of this step is a common reason for the failure of the problem-solving/decision-making process, because ineffective implementation hasn’t been detected and therefore can’t be corrected.<br><br> <li><strong>Evaluate the effectiveness of the Solution.</strong><br> It is particularly important to evaluate the proposed solution in light of the problem statement generated at the beginning of the process. In addition to the expected technical considerations, affective, cognitive, and behavioral outcomes (the impact on the people involved) should be considered.<br><br> <li><strong>Modify the Solution as required by the problems revealed in the evaluation process.</strong><br><br> <li><strong>Repeat the evaluation</strong> cycle to be sure that those problems previously identified and any new ones generated by the changes have all been satisfactorily resolved. <br><br> </ul></ul> <strong>Part 2 - Generating The actual Solution</strong><br><br> What we’ve done up to now has been to confirm that:<br> <ul> <li>We are attempting to solve the right problem <li>Solving it will pay off sufficiently well to be worth the effort expended. <li>We have identified and incorporated all important problem aspects and parameters. </ul><br> These steps are essential to arriving at a viable solution, but are insufficient in and of themselves because they are too general to be implementable. What we must do now to achieve the desired payoff is to actually delve into the specifics of the problem, and then go about developing a detailed solution which will allow us to realize that payoff.<br><br> <strong>CAVEAT!</strong><br><br> The techniques that follow are methods for turning what would otherwise be a completely subjective, basically “gut feel” decision-making process into one which has at least a semblance of objectivity. The problem with it is that, as earlier noted, unlike pure mathematical processes (e.g. addition) which always yield precisely accurate and repeatable results, when a value is generated by one of these methods it is essential to recognize that result is heavily dependent upon the initial assumptions made, and therefore might be quite different, even diametrically reversed, if those initial assumptions are even slightly changed.<br><br> Not only does this mean that it is possible for the analyst to subconsciously influence the decision while maintaining a facade of objectivity, but that any decision reached as a result of such an analysis must be viewed with great caution and even skepticism.<br><br> The ways to counter this problem are to:<br> <ul> <li>discuss and validate the values assigned to the key parameters with other persons who know the situation but have no stake in its outcome. <li>repeat the analysis using a range of each of the subjective input parameters, and then compare the indicated decisions. In this way, you can determine the sensitivity of the analysis to those input assumptions, and then treat the reliability (or non-reliability) of the resulting decisions accordingly. </ul><br> In the examples that follow, for the sake of simplicity in describing these techniques, only one “pass” is made.<br><br> <ul> <li><strong>PMI = Plus/Minus/Interesting of Edward de Bono</strong><br><br> <strong>This is an adaptation of the simple “pros and cons” analysis that has almost universally been used for centuries.</strong><br><br> To apply it: <ul> <li>Construct a table with colimns headed “Plus”, “Minus”, and “Interesting” <li>In the respective columns list the appropriate factors influencing the desired action together with their expected effects <li>Assign a subjective (estimated) positive or negative score for each.effect <li>Add up the scores. </ul><br><br> <strong>A strongly positive score indicates that action is probably justified, a strongly negative score that it should be avoided. In between scores are ambiguous, and allow moving in either direction as desired.</strong><br><br> <center>Example: Deciding whether to relocate from Long Island to New York City.</center><br><br> <div align="center"> <center> <table border="6" width="60%" cellspacing="0" cellpadding="11" bordercolordark="#000000" bordercolorlight="#C0C0C0" bordercolor="#000000" height="373"> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong><font size="4">Plus</strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">Minus</strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">Interesting</strong></td> </tr> <tr> <td align="center" bordercolor="#000000" BORDER="1" height="31">More Social Activity<br>In The City (+5)</td> <td width="20%" align="center" bordercolor="#000000" height="31" BORDER="1">Have To Sell My House (-6)</td> <td bordercolor="#000000" align="center" height="31" BORDER="1">Easier To Find A New Job (+1)</td> </tr> <tr> <td align="center" bordercolor="#000000" height="31" BORDER="1">Much Shorter Commute (+5)</td> <td width="20%" align="center" bordercolor="#000000" height="31" BORDER="1">More Pollution (-3)</td> <td bordercolor="#000000" align="center" height="31" BORDER="1">Meet New People (+2)</td> </tr> <tr> <td align="center" bordercolor="#000000" height="31" BORDER="1">More Interesting Places To Go (+3)</td> <td width="20%" align="center" bordercolor="#000000" height="31" BORDER="1">Less Living Space (-3)</td> <td bordercolor="#000000" align="center" height="31" BORDER="1">More Difficult To Get Work Done (-4)</td> </tr> <tr> <td align="center" bordercolor="#000000" height="31" BORDER="1"> - </td> <td width="20%" align="center" bordercolor="#000000" height="31" BORDER="1">High Rent (-5)</td> <td bordercolor="#000000" align="center" height="31" BORDER="1">More Crime (-3)</td> </tr> <tr> <td align="center" bordercolor="#000000" height="31" BORDER="1">+13</td> <td width="20%" align="center" bordercolor="#000000" height="31" BORDER="1">-17</td> <td bordercolor="#000000" align="center" height="31" BORDER="1">-4</td> </tr> </table> </div> <br><br> <center>Total = 13 -17 - 4 = -8 => Decision: Stay on Long Island.</center><br><br> <li><strong>SWOT ANALYSIS.</strong> In this process you <strong>analyze a situation by identifying its Strengths, Weaknesses, Opportunities, and Threats.</strong><br><br> In its simplest adaptation, a table similar to that for PMI can be created, with 4 columns but in every other way identical. The same problem with the subjectivity of the values assigned to the relevant factors exists, with the same solution. <br><br> The questions that are appropriate for assessing each of the four SWOT factors will vary depending upon the type and context of the problem being analyzed, and will necessarily be very different for (say) a personal problem as opposed to one that is societal, technical, business or political, but certain common themes will exist. <br><br> <strong>Typical questions that will reveal the required information are:</strong><br><br> <ul> <li><strong>Strengths</strong> (always positive) <ul> <li>What assets does your entity (person, organization) possess? <ul> <li>Physical <li>Intellectual <li>Temporal <li>Capital <li>Other </ul> <li>What things does it do well? </ul><br><br> <li><strong>Weaknesses</strong> (always negative) <ul> <li>What assets does your entity lack? <li>What things doesn’t it do well? (i.e. What needs improvement?) <li>What weaknesses do others perceive that you don’t? </ul><br><br> <li><strong>Opportunities</strong> (always positive) <ul> <li>In what immediate and long term ways can your strengths be used effectively? <ul> <li>external factors <li>technological and/or social trends <li>the economy <li>relationships <li>competition <li>etc. </ul> <li>What are the longer range opportunities, and what must be done to take advantage of them? <li>Can you improve the things you don’t do well? </ul><br><br> <li><strong>Threats</strong> (always negative) <ul> <li>What obstacles do you face? <li>What things should be avoided? <li>What is your competition doing? <li>Are the required specifications for your job, products or services changing? <li>Is the current economy, changing societal trends, and/or technology threatening your position and/or prospects? (Are you in an economic bind?) </ul></ul><br><br> The strictly qualitative output of the SWOT analysis can be quantified and analyzed in precisely the same manner as that of both the PMI and Force Field techniques, so they all really represent slightly different (and complementary) ways of achieving the same result.<br><br> <li><strong>FORCE FIELD ANALYSIS</strong><br><br> This method is conceptually very similar to both SWOT and PMI, especially the latter, with the important difference that rather than focusing on factors internal to the success or failure of a proposed solution it describes the external forces arrayed for or against it, so that a decision can be made which takes into account all interests. As with the other methods, this is really nothing more than a specialized and highly refined way of comparing pros and cons.<br><br> <strong>Where a plan has been decided on, Force Field Analysis allows you to appropriately act to reduce the impact of the opposing forces and strengthen/reinforce the supporting forces, as well as to implement changes that may make success more likely .</strong><br><br> <strong>To carry out a Force Field Analysis: <ul> <li>List all forces for change in one column, all forces against in another. <li>Assign a score to each force, from 1 (weak) to 5 (strong). <li>Draw a diagram showing the forces for and against, and their size.</strong><br><br> </ul> <strong>CAUTION!<br><br> The same caveat earlier noted in applying the other techniques regarding the subjectivity of the assigned scores again applies here.</strong><br><br> An example of a Force Field Analysis applied to the same decision regarding whether or not to relocate to the city used earlier in the PMI analysis is shown below:<br><br> <P ALIGN="CENTER"><img src="Force%20Field2.gif" align=middle> </P><br><br> The result of this analysis is only slightly negative, so any decision regarding whether or not to proceed with the proposed move probably shouldn’t be made on the basis of it alone, but rather on the totality of this and any other forms of analysis performed. <br><br> The preferred decision making process consists of a meta-analysis in which each of the various techniques presented herein is applied (as appropriate), and then the final decision summarizes their overall effect. For example, if each of the PMI, SWOT, and Force Field analyses were also slightly negative, then the overall conclusion would be fairly strongly negative. But if one or more was positive, then the overall result would be ambiguous at best. In most cases, because of the considerable impact on the system that would follow from a major change, an ambiguous result will almost always result in continuance of the status quo (= no change). In this problem, both PMI and Force Field analyses are in agreement, so the proposed move to the City is counter-indicated! (But remember again the caveat regarding the subjectivity of the weights assigned to each of the decision criteria!)<br><br> <li><strong>Paired Option Comparison Analysis</strong><br><br> Many real life problems contain several plausible alternative solution options, but little or no objective hard data which will allow you to objectively choose between them. In such cases, a purely subjective technique such as Paired Option Comparison Analysis can still enable the rational selection of the most important problem to solve first, and/or reaching the best solution attainable once that choice has been made.<br><br> <strong>Paired Option Comparison Analysis Procedure: <ul> <li>List the Solution options. (Letter designators are usually assigned for convenience.) <li>Build a table in which each option appears as both a row and a column header. <li>Zero out the table diagonal in which each option is compared with itself, as well as all cells below that diagonal (= the converse duplicates of the above-diagonal cells). <li>Use this table to compare each option with every other option, one-by-one.<br><br> For each such comparison: <ul> <li>Decide which of the two options is the more important. <li>Enter the letter designator of the more important option. <li>Assign a score to show the size of their difference. ( Since this is purely subjective, any convenient scheme may be used: Most common is: 0 = no difference, 1 = small difference, 2 = modest difference, 3 = major difference.) <li>Total the scores for each option. <li>Convert these scores into percentages (if desired). <li>Select the highest scoring option.</strong> </ul></ul><br><br> Example: A local community college student is trying to decide which of the options listed below will be in her best interest to pursue after she receives her Associate’s degree. There’s no good way to accurately quantify their value, but by applying this technique it should still be possible to make a rational decision. <br><br> As she sees the situation, her options are: <ul> <li>Continue on full time to obtain a Bachelor’s degree immediately (A) <li>Obtain some kind of (temporary) employment, while continuing on in night school. (B) <li>Begin a full time career in teaching (C) <li>Use her musical skills to try to “make it” with a Rock band. (D) </ul><br><br> To solve her problem, our hypothetical student creates a Paired Option Comparison Table as follows:<br><br> <div align="center"> <center> <table border="6" width="60%" cellspacing="0" cellpadding="11" bordercolordark="#000000" bordercolorlight="#C0C0C0" bordercolor="#000000" height="373"> <tr> <td align="center" width="30%" colspan="5" align="center" bordercolor="#FFFFFF" height="30"> <strong> <font size="5">Paired Option Comparison Table Example</font></strong></td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong></strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> College (A) </strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong>Night<br>School (B)</strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> Teach (C) </strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> Music (D) </strong></td> </tr> <tr> <td align="center" bordercolor="#000000" BORDER="1" height="31"><strong>College (A)</strong></td> <td width="20%" align="center" bordercolor="#000000" height="31" BORDER="1"><strong></font>-</strong></td> <td bordercolor="#000000" align="center" height="31" BORDER="1"><strong>A,2</strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong>A,1</strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> A,1</strong></td> </tr> <tr> <td align="center" bordercolor="#000000" height="31" BORDER="1"><strong>Night School (B)</strong></td> <td width="20%" align="center" bordercolor="#000000" height="31" BORDER="1"><strong> </font></strong></td> <td bordercolor="#000000" align="center" height="31" BORDER="1"><strong></strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong>C,1</strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> D,2</strong></td> </tr> <tr> <td align="center" bordercolor="#000000" height="31" BORDER="1"><strong>Teach (C)</strong></td> <td width="20%" align="center" bordercolor="#000000" height="31" BORDER="1"><strong> </font></strong></td> <td bordercolor="#000000" align="center" height="31" BORDER="1"><strong></strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> </strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> D,2</strong></td> </tr> <tr> <td align="center" bordercolor="#000000" height="31" BORDER="1"><strong>Music (D)</strong></td> <td width="20%" align="center" bordercolor="#000000" height="31" BORDER="1"><strong> </font></strong></td> <td bordercolor="#000000" align="center" height="31" BORDER="1"><strong></strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong></strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong></strong></td> </tr> </table> </div><br><br> In this case, the analysis has failed to provide a clear decision because our student is equally desirous of continuing on to obtain her Bachelor’s degree immediately or trying to begin a career with a Rock band. But what is painfully apparent is that any thought of attending night school is out, and that teaching is also not really an option worth further consideration!<br><br> This is not the kind of nice, neat answer we would prefer, but the real world is often messy so we shouldn’t really be surprised. But given this ambiguity, the student decides that perhaps the key direct comparison she made between the two dominant options of continuing on in college (A) and trying to establish a career in Rock music (D) should be used as a tie breaker - in effect giving that comparison more weight!<br><br> Checking the table, we see that this direct comparison favored the College (A) option by the slimmest of margins. But we also note that the college option was preferred over each of the other alternatives, so on balance a clear if narrow margin in favor of College (A) emerges as the student’s final decision!<br><br> <li><strong>Grid Analysis</strong><br><br> <strong>This technique is conceptually similar to Paired Option Comparison Analysis, with the important difference that instead of merely directly comparing the opposing options we look at and quantify/evaluate their component factors. So this technique requires much more input information. But in return for that extra input effort, it enables a somewhat less subjective and therefore more precise decision.<br><br> Its major virtue is that it quickly and simply allows the inclusion and evaluation of a wide array of factors while providing a rational decision basis in place of an outright guess or simple “gut feel”, and thus greatly increases the probability that the decision reached will be a good one. </strong><br><br> One interesting fact to be noted here is that the comparisons upon which this technique is based are really the same ones we instinctively use in our normal, everyday decision making. The difference here is that by explicitly listing and quantifying the factors involved we not only assure that all of the relevant ones are included, but that we also evaluate their relative importance correctly. <br><br> <strong>Grid Analysis Procedure: <ul> <li>List the options. <li>List the key factors that will influence the decision. <li>Create a table with options as the row labels, and factors as the column headings. <li>Determine the relative importance of the factors in making the decision (= assign weights to the preferences). If these values aren’t immediately apparent, then a technique like Paired Comparison Analysis may be used to estimate them. <li>Score each option in the table for each of the important factors in your decision. (The usual values assigned are the same 0-3 as before, and it’s not necessary to assign a different score for each option). <li>Multiply each score by its corresponding relative importance value (= weight) to calculate the correct overall decision weight. (The weights assigned usually range from 1-5 for simplicity, but because this comparison is strictly subjective any desired range of values may be used.) <li>Total these weighted scores for each option. <li>As ever, the highest scoring option is the desired choice. </ul><br><br></strong> Example:<br><br> Your family needs a new car, and you want it to have as many of the following desired characteristics as possible: <ul> <li>good acceleration <li>good handling <li>good looks <li>good fully loaded ride <li>V6 engine <li>Lots of cargo room <li>Seats 6 <li>4 wheel drive <li>all power <li>reliable <li>Not too expensive </ul><br><br> The options to be considered are the several major manufacturer’s various models on which the desired characteristics are available. But to keep our table simple for this example we’ll here pretend that only a few manufacturers make suitable models and that each of those manufacturers has only one such model, so that we need only give the manufacturer’s name in our table.<br><br> <div align="center"> <center> <table border="6" width="60%" cellspacing="0" cellpadding="11" bordercolordark="#000000" bordercolorlight="#C0C0C0" bordercolor="#000000" height="373"> <tr> <td align="center" width="100%" colspan="13" align="center" bordercolor="#FFFFFF" height="30"> <strong> <font size="5">Table 1 - Unweighted Factors</font></strong></td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong><font size="4">Factor</strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">Accel</strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">Handl</strong></td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong><font size="4">Appear</strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">Ride</strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">V6</strong></td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong><font size="4">Cargo</strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">Seat</strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">4 Whl</strong></td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong><font size="4">Pwr</strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">Rel</strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">$</strong></td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong><font size="4">Total</strong></td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">Weight</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">Chev</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">1</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">3</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">3</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">1</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">1</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">Dodge</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">1</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">1</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">1</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">Ford</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">1</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">3</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">1</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">Toyota</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">1</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">3</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">3</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">Honda</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">1</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">Nissan</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">1</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> </tr> </table> </div> <br><br> <div align="center"> <center> <table border="6" width="60%" cellspacing="0" cellpadding="11" bordercolordark="#000000" bordercolorlight="#C0C0C0" bordercolor="#000000" height="373"> <tr> <td align="center" width="100%" colspan="13" align="center" bordercolor="#FFFFFF" height="30"> <strong> <font size="5">Table 2 - Weighted Factors</font></strong></td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong><font size="4">Factor</strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">Accel</strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">Handl</strong></td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong><font size="4">Appear</strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">Ride</strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">V6</strong></td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong><font size="4">Cargo</strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">Seat</strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">4 Whl</strong></td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong><font size="4">Pwr</strong></td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">Rel</strong></td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong> <font size="4">$</strong></td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1"><strong><font size="4">Total</strong></td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">Weight</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">3</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">3</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">3</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">3</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">5</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">-</td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">Chev</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">8</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">9</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">5</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">62</td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">Dodge</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">8</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">8</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">3</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">5</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">58</td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">Ford</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">8</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">8</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">3</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">8</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">5</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">64</td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">Toyota</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">8</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">9</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">12</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">10</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">74*</td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">Honda</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">8</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">8</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">2</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">8</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">10</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">68</td> </tr> <tr> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">Nissan</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">8</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">8</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">3</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">4</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">6</td> <td width="20%" align="center" bordercolor="#000000" height="20" BORDER="1">8</td> <td width="34%" align="center" bordercolor="#000000" height="20" BORDER="1">10</td> <td width="33%" align="center" bordercolor="#000000" height="20" BORDER="1">67</td> </tr> </table> </div> <br><br> From Table 2, the preferred decision is immediately obvious! The Toyota clearly outscores all of the others. <br><br> <strong>CAUTION!</strong><br><br> As earlier noted, the key thing to be wary of here is that this is a strictly subjective process which represents the preferences of the individual assessing and specifying both the factors to be included and the weights assigned to each. So each and every individual conducting precisely this same analysis will almost necessarily obtain different totals, and thus may reach quite different decisions!<br><br> <li><strong>Decision Trees</strong><br><br> The relatively simple techniques we’ve just explored are applicable in a very wide set of circumstances, but unfortunately they’re not sufficient to handle situations in which complex financial or similar numeric calculations are required to adequately define the problem and its solution. That requires the more powerful and complex tool of Decision Trees, which we discuss next.<br><br> Decision trees are an excellent tool for making financial/number based decisions because they provide a simple visual framework within which:<br><br> <ul> <li>All choices can be viewed, discussed and challenged. <li>Alternative decisions and the implications of taking each can be spelled out and evaluated. <li>The value of outcomes and the probability of achieving them can be quantified. <li>An accurate, balanced picture of the risks and rewards that can result from each alternative decision can be derived. <li>The best possible decisions can be objectively made, given the existing information and best available estimates. </ul></ul><br><br> How to (Manually) Draw a Decision Tree<br><br> <ul> Represent the decision that needs to be made by a small square at the left edge of a large sheet of paper. From this node: <ul> <li>Draw a line toward the right for each possible solution, keeping them as far apart as possible to allow for later expansion of your thoughts. <li>Briefly name/describe each solution along its line. <li>At the end of each solution line, consider the results of taking that decision: <ul> <li>If uncertain, draw a small circle. <li>If the result is another decision that needs to be made, draw another square. </ul> <li>Specify the decision to be made or the factor to be considered above its square or circle. <li>If you have completed the overall Problem Solution at the end of any line, just leave its termination blank. <li>Starting from each new decision square on your diagram, again draw lines representing the options that could be taken. <ul> <li>From the circles draw lines representing possible outcomes. <li>In each case, again enter a brief explanatory note on the corresponding line. <li>Repeat this process until you have included all of the possible outcomes and decisions you can visualize resulting from your original decision. </ul><br><br> This is only the first step in the process!<br><br> <li>Review the Decision Tree diagram and examine/question each square and circle to be sure that you haven’t missed any solutions or outcomes that should be included. If you discover any such, draw them in and redraft the tree as necessary (if parts of it have become too congested or untidy for clarity).<br><br> You should now have a good understanding of the range of possible outcomes.<br><br></ul> <li>Making a Preliminary Evaluation From Your Decision Tree<br><br> In this step, you perform some simple calculations to reveal the decision that has the greatest payoff:<br><br> <ul> <li>Assign a cash or numeric value to each possible outcome, representing how much you think it would be worth to you. (Caution! This is another subjective judgment which will strongly affect your final decision, so these values must be chosen with extreme care!) <li>For each circle (representing an uncertainty point), estimate the probability of each outcome. If you have data on past events you may be able to make fairly rigorous estimates of the probabilities. Otherwise write down your best guesstimates. (These too will be subjective judgments, so the same caveat applies!) <ul> <li>Probabilities must total 1.0 <li>Percentages must total 100%. </ul></ul><br><br> The result will be a tree like the one below:<br><br> <P ALIGN="CENTER"><img src="Decision%20Tree3.gif" align=middle> </P><br><br> <li><strong>Calculating Tree Values</strong><br><br> Once you have quantified the outcomes and assessed the probabilities of each, it is time to start calculating the values on which to base your decision: <ul> <li>Start on the right hand side of the decision tree, and work back towards the left. </ul><br><br> <li><strong>Calculating Values For Uncertain Outcomes (circles on the tree diagram)</strong><br><br> <ul> <li>Multiply the value of each outcome by its probability, and note the result. <ul> <li>The total value of any node of the tree is gained by adding together the values for its branches .(Remember that the sum of the probabilities assigned to each node’s branches must total 1.0) <li>As the set of calculations for each node (decision square or uncertainty circle) is completed, simply record that result in the appropriate place on the right edge of the tree. <li>After this, all the calculations that led to that result can be ignored! (i.e. that branch of the tree can be discarded.) This is called 'pruning the tree'. </ul><br><br> In the example above, the value for 'College, Ph.D.' is:<br><br> 0.2 (probability, best outcome) x $450,000 (value) = $ 90,000<br> 0.6 (probability, expected outcome) x $250,000 (value) = $75,000<br> 0.2 (probability, worst outcome) x $200,000 (value) = $ 40,000<br> -------------------------------------<br> Node Total = $205,000<br><br> These values are then entered in alongside their respective nodes as shown below:<br><br> <P ALIGN="CENTER"><img src="Decision%20Tree5.gif" align=middle> </P><br><br> <li><strong>Calculating The Value of The Decision Nodes</strong><br><br> <ul> <li>Redraw the tree, eliminating all of the branches of the uncertain nodes and replacing each with the corresponding single node value just calculated. </ul> <ul> <li>Enter the cost of each option along the corresponding decision line. <li>Subtract the cost from the node value just calculated, to find the benefit of that decision. “Sunk costs”( = amounts already spent), are not considered in this analysis. </ul><br><br> Calculation of the decision (“choice”) nodes in our example is shown next:<br><br> <P ALIGN="CENTER"><img src="Decision%20Tree6.gif" align=middle> </P><br><br> <li><strong>Arriving At The Decision</strong><br><br> When you have calculated the net benefit of each alternative decision, simply select the one which has the largest value.<br><br> In this example, the benefit we calculated for “College, Ph.D.” was $205,000 and its cost was $180,000, giving a net benefit of $25,000. But the benefit of “College, B.S.” was $280,000. On this branch we therefore choose that as the most valuable option, and allocate its value of $280,000 to the decision node for the “college” branch of our decision tree.<br><br> Similarly, we calculate the values for the two branches of the “Work” approach as $380,000 and $370,000, so of these the “use music skills” option is slightly better. And overall it’s equally clear that it’s also best, so it and its net benefit of $380,000 become our final decision and value.<br><br> <strong>Caution!</strong><br><br> Our application of the Decision Tree technique in this situation produced the indicated decision that it would be best for this hypothetical student to go to work immediately after high school rather than to college, and that utilizing his music skills would be only marginally more productive (in terms of economic payoff) than not doing so.<br><br> But that decision was heavily dependent upon both the 10 year term chosen for the evaluation, and the probabilities assigned to the various outcomes, for which, unlike ones made in a business context where it’s often possible to use prior history to obtain somewhat more reliable estimates of the probabilities involved, in this context there is no realistic way of sharpening those estimates. So the only feasible approach which can improve the probability that the decision reached is at least reasonably valid is to repeat the analysis using somewhat different probability estimates for the various outcomes, and using at least one other, substantially longer payoff period (e.g. 20 years in place of 10). Then combining this form of analysis with at least some of the others discussed to produce a meta-analysis is also wise.<br><br> In this situation that’s especially necessary, because the application of this quantitatively oriented technique to a problem in which the most important payoff for many people is qualitative (e.g. quality of life - satisfaction with what you’re doing and with the people with whom that forces you to be in daily contact) rather than quantitative (the amount of money you earn), can very easily produce a ridiculous result. The problem that the incautious analyst encounters is that the mechanical application of the technique produces a nice, neat numerical result which has every appearance of both accuracy and validity, when in many cases it really possesses neither!<br><br> So, as with any other technique, Decision Tree Analysis should always be applied in conjunction with a liberal helping of good old fashioned common sense!<br><br> <br><br><br><br> <strong>Click Here To Move On To</strong><a href="RDMCh6.html"><font size=+1><font Color="#0033FF"><strong> Chapter 6</strong></font></a> <br><br> <strong>Click Here To Return To</strong><a href="RDMCh4.html"><font size=+1><font Color="#0033FF"><strong> Chapter 4 </strong></font></a> <br><br> <strong>Click Here To Return To</strong><a href="RDMTOC.html"><font size=+1><font Color="#0033FF"><strong> Table Of Contents </strong></font></a> <br><br> <strong>Click Here To Return To</strong><a href="index.html"><font size=+1><font Color="#0033FF"><strong> Milt's Go Page</strong></font></a> <br> <br><br> <strong>Click Here To Email Your Comments/Suggestions To</strong><font size=+2><font color="#0033FF"><a href="mailto:bradleym@eniinternet.com?subject=Reasoning And Decision Making Comments/Suggestions"> Milton N. Bradley</font></a> </td> <td width="100"> </td> </tr> </table> </body> <!-- Mirrored from users.eniinternet.com/bradleym/RDMCh5.html by HTTrack Website Copier/3.x [XR&CO'2014], Sun, 06 Nov 2022 06:59:39 GMT --> </html>