Parts of the Management Process – Test Questions

Parts of the Managing Process - Test Questions

Marks Available – Figures in brackets at the end of the question show the marks available for each question,

1 Write the meaning of each of (a) Planning (b) Directing and (c) Checking in simple words (without using the words you have to explain) . Start: each meaning with a verb. (3+1=4)

2 Which one of the three parts of the management process (as defined in the notes) do you believe you carry out least effectively. In one sentence give a reason (or reasons) for your belief that this particular one rates as worse than the other two. in one sentence state what you plan to do to improve your effectiveness of the least-effective part. (2+24)

3 Explain, and give an example to show, “the Parts of the Managing Process do not necessarily constitute a continuous process for any one objective”. (3)

4 Give one example for each of (a) conscious planning and  (b) unconscious planning (2+l=3)

5 You hold a management position and determine to do a job  which will involve at least six of your Subordinates. You sit down by yourself to plan to carry out the job. List five heading sidentifying areas or aspects of the job which probably would help you plan the particular job better. (This question does not ask you to list the stages of planning.) (5+1-6)

6 In planning to achieve an objective, Planners need to make a number of decisions and implement them if the plan will achieve its end objective. However some plans will include decisions on some things which the Planner knows (while planning) may rate as unnecessary. In one sentence explain these type of decisions. (2)

7 Give one example of a decision that does fall within the class of objective setting. (2)

8 Give one reason to show why you agree, or disagree, with  the following statement (as referring to yourself) – “I would prefer to define decision making in my organisation as an activity which included both making, and implementing, decisions. (3)

9 Give one reason why you agree, or disagree, with the following statement – “Trying to influence another person must involve trying to communicate with the person (2)

10 Give one example for each of the following four types of  Directing Techniques – (a) Order, (b) Advice, (c) Information Giving, and (d) Information Seeking. (4+1=5)

11 Give an example of a question which most people in your organisation would accept as an order. (2)

12 Describe one actual situation in which it would prove better for a Manager to give an indirect’ direction (as opposed to a direct” direction) to a Subordinate. Give one reason why you would rate the ‘indirect’ direction as more suitable for the situation you describe, as compared with a “direct” direction. Do NOT use the case of a Subordinate who needs an indirect direction because of his/her personality. (4)

13 Give one reason (and use an example) to explain why you agree, or disagree, with the statement “In some situations the measurement and evaluation of an activity often happens simultaneously”. (4)

14 Use the word “check” in two sentences. In the first  sentence the word should mean “measure” and the sentence makes it clear that the check does NOT include the aspect of
compare the measurement with a standard. (2+2=4)

15 Give one reason to show why the following statement does, or does not, apply to your own working approach – ‘Many Managers do not consciously think about checking and/or spend too little time on it”. (3)

16 Give two examples of a standard which your Organisation  uses in checking in the area in which you work. (2+1=3)

17 Explain, and give an example, to show that all standards do not rate as objectives. (3)

18 Explain the meaning of the statement – “A useful standard would reflect progress toward an objective”. Include in your explanation what would occur if someone used a standard which did not reflect (accurately) progress toward an objective. (2+2=4)

19 Give one example which will show that “checking alone will not control anything”. (3)

20 Give a standard which your organisation uses which you  rate as wrong to use. In one sentence, state why you believe your organisation should not use the standard. (3)

21 Give one reason to explain why you agree or disagree with the statement – People who carry out the activity of managing have more flexibility than if they carry out the activity of controlling”. (3)

22 Give one question you could ask which would allow you to apply in practice the difference between Managing and Controlling – as given in the notes. (3)

23 Give one example which shows that a suitable objective yesterday’ has become an unsuitable one today.  (3)

24 Give one example for each of the following sub-sections  to explain how –

     (a) Poor checking lead to a small and inadequate amount of coordination.

(b) Good planning lead to a large and useful amount of coordination. (2+24)
 

Drawing Decision Trees

Introduction

1 These notes aim to help people draw Decision Trees. They describe a Decision Tree and illustrate it. The notes also list what many people do when they make decisions and point out that a Decision Tree uses most of the ideas in a written and systematic way.

2 The notes discuss a detailed step-by-step approach to drawing Decision Trees and discuss a number of general points that should help the Reader to “grow better Trees”.

 

Decision Making & Objective Setting

3 These notes define decision making as – a conscious thinking process aimed at choosing between two or more things. A particular class of decision making involves objective setting: a conscious thinking process aimed at choosing between two or more objectives.

4 Planning involves two stages: (a) find objectives and (b) choose between objectives, and objective setting fits into the second stage of planning.

5 Decision Trees aim not only to help make decisions (choose objectives) but also to help find possibilities. Thus, it provides a planning technique that aims to help both stages

A Description of a Decision Tree

6 A Decision Tree provides a method of depicting possible decisions and their possible outcomes in a logical arrangement and sequence. The following diagram shows one way of drawing a Decision Tree –

 

 

 

 

 

 

7 Decision Trees aim to group like factors together and to display possibilities in a logical and written arrangement. Usually they start with different branches (one for each possible decision). Each of these decision branches ends in a node. From each node come branches that show possible outcomes/results (actions by others) and/or consequences.

8 The outcomes of decisions include competitive moves as well as chance factors.

9 Example. A decision to market an improved product at a higher price may lead one or more opposition companies to decrease the price for their unimproved product.

10 A Tree aims to show “every” possibility at each decision point. However, this approach could mean a very large number of branches from some points. To avoid this situation occurring, people can group like possibilities together, until the number becomes manageable.

11 Planners can label one branch as “other”. In general, Planners should avoid this approach since planners use a Decision Tree to help search for additional answers to a problem. The use of “other” makes it simple to stop searching.

12 While some articles use the sloping branches approach used in the above Tree, the “square” approach used in later diagrams allows people to fit more detail into a given area.
What people normally do when they make decisions as compared with the drawing of a decision tree

13 People make many decisions within their own mind without ever writing things down or trying to draw any type of structure such as a Decision Tree. A consideration of the various things people do will help review the desirability, or otherwise, of using Decision Trees.

14 People often do not separate clearly the two major stages of planning (find objectives and choose between them). Sometimes this approach leads people to choose between possibilities before they give adequate time to the process of finding further possibilities.

15 Experienced people often use this distinction in their own thinking. Decision Trees help to make thetask of finding possibilities a separate stage. Thus, the technique helps to keep people working on – find possibilities, i.e. people avoid the trap of moving too quickly into the stage of choosing between possibilities.

16 Many people find it useful to write down possibilities. This approach means they cannot forget them temporarily and reading over the list often brings others to mind. Such a list also helps to separate possible actions from possible results.

17 Some classification of the possibilities in the list by grouping things that have some common elements will often help to find further possibilities. It may also help to evaluate some of the possibilities.

18 Another approach aims to evaluate possibilities more systematically. One method states: select a number of the likely possibilities and write down the advantages and disadvantages of each. This approach formalises what people often do in their own mind.

19 Sometimes people gain help in finding more possibilities by seeking someone else’s opinion. This approach means that one person must communicate ideas to another person. To help do so and make better use of the Adviser’s time, some people find it better to write down the points they have in their mind in some systematic and logical way.

20 With most complex problems, people find difficulties keeping in their head the many factor involved and their interrelationships. It does not prove easy to take each likely possibility and think about likely results. Some method of writing the material down soon becomes essential if people want to explore systematically the possibilities.

21 People need help to think more clearly about the possibilities of various results occurring. Most people do so in their thinking when they classify various possibilities as “very likely” or “unlikely” etc. However, these terms rate as rather loose and people should use more specific quantitative terms. Instead of rating the probability of outcome “X” occurring as “very high” or “very low” or “about average” etc., people should use quantitative terms with reference to factors divided into 100 parts.

22 Thus if something has a very low probability, people can say it has one chance in 100 of occurring. In this case, the number 0.01 describes the probability. The probability number 0.50 refers to something that has a 50/50 chance of occurring. Something with a very high probability of occurring would use the figure of 0.95 or 0.99.

23 Example A probability of 0.99 might occur when one action involves sacking someone. In some cases, everybody in Management might believe that a l00 probability exists that a strike will occur as a consequence of the sacking.

24 All the above things can occur in a person’s mind but this approach has disadvantages. First, forgetting takes place. Second, often people only see patterns and relationships between factors when they write them down.

25 The above section aims to show that a Decision Tree does not rate as very different from what happens in real life – for important decisions.

 

Why do people (a) have difficulty in drawing, and (b) avoid using decision trees?

26 Often people find difficulties in drawing Decision Trees. However, they would find that if they considered the same problem in their own minds they would have just as much difficulty.

27 In actual practice, they might appear to have less difficulty in using their mind. However probably this situation happens because they exclude some of the factors and their interrelationships from consideration.

28 Decision Trees rate as a useful tool for Executives. However, they encourage people to spend moretime on thinking. However, thinking involves hard work and therefore people often try to avoid it. In addition, people get easily distracted from thinking about the one problem. Further, they always have the overall excuse: “I haven’t got time”.

29 To discipline oneself to draw a Decision Tree and look at all the possible outcomes will not prove easy. In any case, it delays the actual making of a decision and most people find it very satisfying and relieving (in the short term) that they have made a decision. It means that they do not have to worry about the problem any longer. Most people put aside the possibility that they have made a wrong decision and they could have many more worries in the future. The temptation always exists to risk that their decision will not prove wrong.

30 Further, it will prove difficult to depict some situations in the form of a Decision Tree. Sometimes some type of table or graph will provide a better arrangement of some types of information involved in the decision- making process.

31 In addition, people need some skill in drawing Decision Trees and disciplining their mind to think in a particular way. People will find that drawing their first ten or twenty Decision Trees will prove more difficult and a lot more time consuming than drawing Trees when relatively skilled.

32 Thus people need to gain experience and practice in drawing Decision Trees in order to make it likely they will use them.

33 Initial attempts to draw a Decision Tree often prove rather difficult because the selection of all the possibilities and their logical grouping will prove difficult. The following points which should help in this task.

 

Some General Points

34 Before starting a Tree, Planners should identify clearly the problem or decision under consideration and, preferably, put it in writing.

35 In the early stages Planners should distinguish clearly between (a) possibilities and (b) the evaluation of them (i.e. possible results) Thus, initially, Planners should just identify possible objectives without trying to evaluate them. Evaluation often encourages people not to record possibilities that sound “bad” or “silly”. Since the Tree approach aims to record all possibilities, such an early evaluation would reduce the chance of achieving this aim.

 

A suggested procedure for finding Branches for Decision Trees

36 The steps listed below apply to finding branches for both possible objectives and possible results. In future, these notes use the word “possibilities” to apply to both objectives and results. However initially they emphasise possible objectives. Most examples refer to objectives –

37 After the initial listing of the steps, the notes discuss each step and give an example to illustrate how to use each step.

38 The steps do not necessarily follow strictly one after the other. People can use any of the steps for thinking of objectives at different times during the “growing” of a Tree.

39 Step 1 – List possibilities in writing.

40 Step 2. Look for, and list: (a) any common element found in two or more of the possibilities and/or (b) any important elements which rate as part of one of the possibilities (or common elements) listed. Consider the “correctness” of the major words used for the purpose of the Tree.

41 Step 3 Look for, and record, “logical opposites” to either possibilities, part-possibilities, or common elements. (Later sections explain the meaning of logical opposites.)

42 Step 4. Re-write the possibilities in some type of Tree form, putting like things together.

43 Step 5 Take each node and consider the branches recorded that follow it. Try to define all the branches that will describe completely all possibilities from the particular node or fork of the branch. Re-draw the Tree, if necessary.

44 Step 6. Use a branch called “other” where (a) you have not found all possibilities from the one node and (b) spending further time at that stage does not seem worthwhile.

45 Step 7. Consider the start of the Decision Tree and try “moving to the left”, i.e. find the “logical opposites” to the first branch.[1]

46 Step 8. Repeat Steps 2 to 6 and record any new possibilities found (i.e. Step 1).

47 It proves difficult to know just how long to continue this process. It will become a matter of judgement by Planners. Probably it should depend on the (apparent) completeness of the Tree and the importance of the end objective for which the Planner wants to draw the Tree.

48 The following paragraphs give more details on the steps described above.

Step 1: List possibilities in writing

49 Planners can (and should) carry out this step at any time during the drawing of a Tree. People forget possibilities quite easily; thus if a possibility comes to mind, record it.

50 Planners do not necessarily have to put any possibility in the correct place on the Tree: Planners who cannot find the right place easily should record it on the side of the page or on a separate page.

51 After Planners use each possibility (not recorded on the Tree), they should show themselves that they have done so by drawing a line through it. Some such systematic procedure should ensure that they have used all possibilities on the Tree or they have discarded them.

52 Example. To make these steps understandable the following sections include an actual example. Each step will show some aspect of the example.

53 The example has the objective “Increase sales of Product X’”. Step 1 produces the following result:

(a) Give Sales Representatives more training.

(b) Look for more Distributors to take on the product.

(c) Put on more Sales Representatives.

(d) Improve the product.

(e) Drop the price.

(f) Improve the package.

(g) Alter the Product’s name.

(h) Find other uses for the product.

(i) Collect opinions from people who buy it and find out why they buy it.

54 Many more possibilities exist but the above provides sufficient to make the example useful.

55 Step 2: Look for, and list: (a) any common element found in two or more of the possibilities and/or (b) any important elements which rate as part of one of the possibilities (or common elements) listed. Consider the “correctness” of the major words used for the purpose of the Tree.

56 One approach to finding a common element involves looking for possibilities that include the same word.

57 Example. In the above example, two possibilities include the term Sales Representatives. Another common element exists - the first three possibilities and the last one all relate to people. The other possibilities all relate to something other than people.

58 Another common element exists between “Improve the product” and “Find other uses for it”. Both relate to how people use the product.

59 Product and Package have a common element in that they both refer to physical aspects of the products. Similarly, Name and Price both refer to intangible aspects.

60 The example proves just as fruitful in looking for important elements in the original possibilities listed and some of the added later ones.

61 Example. The common element “people” contains important part elements. For example, people can include the classes Buyers or Sellers. (The Sellers include both Sales Representatives and Distributors, listed in the possibilities.) To become more useful, Planners could use the classification Buyers of Product “X” and Sellers of Product “X”. However, people have other classifications. A number of people exist who do not buy Product “X”. This thought produces a better classification – Sellers and Non-Sellers and Buyers and Non-Buyers.

62 This idea leads to the consideration of whether people exist who could sell the product and at present do not. Similarly, do people exist who rate as Non-Buyers and could become Buyers? (This point rates as the central point of the whole problem.)

63 if Planners consider the possibility of collecting opinions from people who have bought “X”, logically they should look at the opposite[2] of this group, i.e. people who have not bought the product. Planners can easily move down the sub-classification to (a) the non-buyers who considered buying it and (h) those non- buyers who have never considered buying it. Those who have never considered buying “X” divide into those who have (a) heard of the product and (b) not heard of the product.

 

 

64 Further thoughts indicate that buyers of the product also have two sub-classes: Buyers of our brand of Product and Buyers of other brands of the Product.

65 The following diagram shows the above possibilities.

Possible Actions to Take Concerning Various Factors \

Diagram 1

Diagram 2

People

 

 

Diagram 3

 

66 At this stage, the above diagrams show three separate Trees. It seems a better plan to approach the problem this way, i.e. record what does fit together quite readily and then look for combinations.

67 Example. Planners can easily combine diagrams (I) and (2) by tacking (I) on to the top line of (2). The use of diagram (3) will mean a re-writing of the top part of Tree (2). A later diagram does so.

Step 3: Look for, and record, “logical opposites” to possibilities, part-possibilities, or common elements.

68 The words “logical opposite” mean more than the opposite of something, although this term includes ordinary opposites.

69 Examples. “Up” rates as the opposite to “down”, However if Planners seek “logical opposites” then going sideways would qualify, even though “side” does not rate as the ordinary opposite of “up”. The “logical opposite” (and opposite) of “move to the left” would say “move to the right”, but “move forward” or “move back” would rate as “logical opposites” as well. In addition, the logical opposite of “remain still” exists.

70 Finding “logical opposites” requires some imagination but with a little experience and practice the idea will prove easy to use. Further, many situations exist in which common “logical opposites” often occur. These opposites will help to draw a Tree for such situations.

 

 

71 Example. In the example, consider the possibility of - add to the number of Sales Representatives (or Distributors). The common or universal “logical opposites” rate as -
“increase”, “no change”, and “decrease” – with respect to the number of people concerned. The same “logical opposites” apply to price.

72 In considering “Find other uses for the Product”, the two-part classification comes to mind – Present use, Possible uses. Possible uses means future uses. This stage does not include the “logical opposite” – consider past uses which today’s consumer do not use. Probably this idea will not prove very useful but it could happen that a Company could try to revitalise a forgotten past use (perhaps after amendment) and gain acceptance (e.g. when should someone next try to re-introduce the Hula Hoop or the Yo-yo?)

 

Step 4: Re-write the possibilities in some type of tree form – putting like things together.

73 The Tree on the next page rates as appropriate. Readers should understand that this topic equals a very large one (and a large Tree) and it does not prove easy to draw.

74 The Tree on the next page merely sets out various possibilities in one systematic and logical manner. It does not specify what action to take, simply that someone could take some action with the various items identified. The specific aspects identified such as “increase our Distributors or our Sales Representatives or the Price” rate as the only specific actions listed. The rest merely identify areas.

75 The actual things that someone could do in the specific areas listed need consideration. The following possibility (already identified) exists: “collect opinions from people who buy the product and find out why they buy it”. The above analysis has shown that a number of other people exist from whom it may prove worthwhile to collect opinions. It may prove desirable to draw a Tree for this particular area. At all events, it will provide another example.

76 In the actual diagram above, the four branches: “Non-Sellers”, “Sellers of Product ‘X’”, “Buyers of Product ‘Y’”, and “Non-Buyers of Product ‘X’” do not rate as mutually exclusive. Someone could fit into the classes of both a non-seller of Product “X” and a buyer. A number of other pairs such as this pair exist. These notes ignore this problem; but will consider the actual problem of looking at buyers from whom to collect opinions. The first attempt at looking at the aspect more logically gives the following diagram.

 

 

 

 

 

 

 

 

Diagram 4

Take

Action

With

Res-

pect

to -

Diagram 5

77 However even the above does not rate as completely consistent because a buyer of Product “X” who has never heard of Product “X” cannot exist. Therefore, the Tree needs re-drawing – as occurs later.

78 At this stage, an example of a different approach should prove useful. Sometimes a diagram will help to clarify a person’s thinking better than a Tree. The following diagram aimed to help see what possibilities exist. It shows the relationship between (a) Buyers/Non- Buyers and (b) people who have considered the various brands of the products or have not considered one or more of them. The central part of the diagram shows whether the combination can, or cannot, occur.

Diagram 6

 

Considered Brand Bought

Considered Brand Bought and Others

Never Considered the Product

Buyers of Product “X” Brand “A”

Must occur

Possible

Impossible

Buyers of Product “X” Other Brands

Must occur

Possible

Impossible

Never bought Product “X”

Impossible[3]

Impossible3

Possible

 

79 With Diagram 6 completed it proves easier to re-draw the Tree to get a classification which does not overlap in its categories. Diagram 7 shows this tree.

Diagram 7

 

80 Even the above diagram does not allow for “Buyer of both Product “X” and other brands”.

81 A Planner could add more detail to the above diagram in the area marked (A) by using the following:

Diagram 8

 

 

 

82 A Planner could sub divide the branch “Never bought Product “X” in a different way as shown in the following diagram

 Diagram 9

83 These notes have not defined the word “Buyer” as used in the diagram above. A Planner would introduce a further complexity by talking about various levels of buying strength.

84 Example. Regular buyers over the last two years, buyers who have just started buying the product, buyers who have just changed from one brand to our brand, and so on. The possible complexities will become almost infinite. The desirability of taking them to this extreme would depend on the activities involved and the importance of the whole objective in this area in particular.

85 At this stage Readers may well ask – Just where will this approach lead? and Will it prove useful?

86 First, it does identify the people from whom investigators can seek opinions. However – could a Planner have reached the classifications in the latter diagrams and understood some of their relationships by using a better/shorter method? Would a less-complete classification prove just as useful?

87 The next stage would become – decide what opinions to collect. The following question shows the scope: “Why did you buy, or did not buy, the particular product and/or particular brand of the product?”

88 Once someone has identified the classifications then someone can aim to evaluate the possible results of the various types of questions to the various classes of people. Perhaps researchers will not need opinions in a particular market research from people who have never considered buying the product. Early in the interview once an Interviewer identifies a Respondent as belonging to this particular category, they would ask no more questions. It can also occur that certain types of questions will prove useless to ask of people who have never considered “our Brand”.

89 With some grouping of these classes of people and the addition of why they did (or did not) buy, it will prove easier to look at the possibilities of various marketing activities having different effects.

90 Example Investgators could look at the activity of altering the price from the viewpoint of these various classes of people. They could try to assess the likelihood of them buying more, less, or having little effect - in relation to a price increase, or decrease or no change.

91 The possibility “Collect opinions, etc.” has proved an extremely complex one. However, this type of situation often happens in an organisation: someone asks someone to do something that rates as very broad and unless someone works out the details often what people do proves wasted. Quite often, the actual amount of work exceeds what the original person suggesting the approach envisaged. Sometimes people carry out the broad objective quite poorly. People should not feel surprised at the useless results that often occur.

 

Step 7: Consider the start of the Decision Tree and try “moving to the left”, i.e. find the “logical opposites” to the first branch.

92 The discussion of the example has illustrated the processes of Step 5 and Step 6 but it would take a great deal more effort to draw the whole Tree for the objective of the example and it would cover a very large sheet of paper.

93 Example. It will prove simple to find some “logical opposites” of the objective - “Increase sales of Product ‘X’”, The possibilities include: (a) maintain the present sales, (b) decrease the sales, or (c) increase them. The area of decreasing sales has an end to it in the sense that the objective could state – achieve no sales. This objective equals deleting the product from the company’s operations. Although Planners may have considered this objective, they should keep it well to the fore since it will prove easy to forget or overlook this “obvious” possibility.

94 Investigators could first aim to establish the likelihood that sales of the product can increase. They can compare two objectives – (a) spend money for market research on finding out why people buy and do not buy the product versus (b) ask questions about the real need which the product has in the past satisfied and will anything change (reduce) this need and will other products satisfy it better,

General Points

95 Apart from the steps discussed above the following general points should help in drawing Decision Trees.

96 Check that you aim to draw only one Tree at a time. Sometimes a number of problems will exist at the same time and interact. A Tree may prove difficult to draw because no one has established the separating point between the problems. Planners should do so.

97 Do not try too hard to fit all the possibilities into the initial Tree. Planners who have difficulty finding a place for a possibility in a Tree should write it down on one side of the page. At a later stage, it may prove much easier to fit it into the structure – because the drawing of the Tree has established a better understanding of the whole situation. Eventually it must prove possible to fit it in or feel sure that it belongs to a different problem.

98 With complex problems, it will often help to draw “small” Trees for different parts of the problem and then try to put them together.

99 Develop the Tree from all nodes on each vertical level – rather than extending the Tree on one horizontal level much further than at other levels.

100 In drawing the possibilities from each node, group together similar possibilities. Probably no node should have more than about five or six branches coming from it.

101 Use the largest concept or idea in deciding on a branch.

102 Examples. Profitability rates as a larger concept that sales since the interaction between sales and expenses gives profit. “Go home” (for a person already at work) gives a “smaller” idea than “Leave work”. In this case “Go home” rates as the smaller concept since “Go home” means that the person must “Leave work”, but the person could also “Go to the races” or “Go out to lunch”. The larger concept “Leave work” covers all the possibilities but “Leave work” does not necessarily mean that the person must “Go home”.

103 Decide when another decision node(as opposed to a chance node) should follow an originaldecision node.Similarly decide when a chance node should follow another chance node.

104 Some two-part decisions only help to group possibilities logically. Thus some things to do must follow a “do something” (versus “do nothing”) node.

105 A variation of this point states – some possible decisions have a sequential link; they can never exist as alternatives.

106 Example. Probably the objectives - “Phone the Supervisor” and “Ignore the different instructions given” rate as sequential. The following diagram shows an appropriate Tree of the possibilities.

Diagram 10.

107 A slightly-different form might improve the tree:

Diagram 11

108 The diagram in a different form shows more clearly that the action of informing someone rates as independent of the decision to accept or reject different instructions.

In the context of - I will inform my Supervisor, I will not carry out (i.e. I reject) his instructions, the term ignore does not fit. Usually people ignore instructions but do not tell their Supervisor that they intend to ignore the instructions. The term “passive resistance” describes this situation

109 Another reason exists which supports the idea that a chance factor should not necessarily follow a decision node. Sometimes the particular methods used to implement a decision will prove almost as significant as the decision. In this case, the reactions to the decision should really take into account the means of implementing the decision.

110 This point holds true particularly in personnel problems. However with broader scale objectives (e.g. building a new factory or launch a new product) the implementation rates as important, but often it will prove useful to assume that the implementation will proceed with average efficiency. Where implementation largely depends on one or a few factors (e.g. the persuasiveness of one person) then Planners should link the methods with the implementation of the decision.

111 Look for critical and significant possibilitieswhen grouping and arranging possibilities.

112 Example. In looking at consequences of a decision, a critical point often becomes whether the result will reach a break-even point (where income equals expenditure). Thus, the three branches from a consequence or chance node might show (a) makes a profit, (b) breaks even, (c) makes a loss.

113 One interesting example occurs with the result – “effect on productivity”. It can involve increase, no change, and decrease. However, sometimes Planners should look carefully at a special point of decrease, i.e. nil productivity. In certain industrial situations, this point means that a strike has resulted.

114 Recording branches in order on some scale.

115 Examples. Record good, bad, or neutral results in order: Good, neutral, bad (or bad, neutral, good). High Profit, Medium Profit, Break Even, Loss shows another set of possibilities arranged in order

 116 Sometimes Planners should put all the good possibilities at a higher (branch) level than poor ones. This point assumes that Planners can evaluate easily the “goodness” of a branch. In some cases, it can (e.g. profit versus loss); in others it cannot. (This suggestion also tends to oppose the previous suggestion of not evaluating possibilities but sometimes the evaluation rates as obvious and not a matter of personal opinion.)

117 Remember that Planners draw a Tree to help make decisions. The provision of many branches or their elaboration can waste time and money. However once someone makes particular decisions then it will prove appropriate to elaborate (draw another Tree) on the particular branches chosen.

 

Universal Trees

118 Some Trees exist which rate as appropriate for a number of problems. Their structure may need adaptation for particular problems; they will certainly need translation.

 

 

 

 

 

 

 

 

119 The following shows one example.

“Universal” Decision Tree Form for Decisions

120 This Tree could not show easily all the possible consequences that might arise from any one decision since a large number of decisions exist. However, Planners could select the four or six “best” and draw the consequences of these decisions.

121 For Planners who select the decision “collect further information”, methods of continuing or abandoning the objective will not prove appropriate until someone obtains the “further information”. One method of continuing the Tree would involve listing the likely broad results that might come from the further investigation. Then it could show decisions on how to continue or abandon. However probably this approach would prove inappropriate unless the whole matter rated as so urgent that Planners should spend time in planning to enable a quick start on continuing the problem immediately they obtained “further information”. Unless this situation existed, it would prove better to defer further thinking until someone obtained the “further information”.

 

Summary

122 A Decision Tree shows possible decisions and their possible outcomes written out in a logical arrangement and sequence. It aims to help both stages of planning, i.e. (a) find possibilities (objectives) and (b) choose between possibilities. These notes call the second stage of planning “decision making”.

123 In Decision Trees, Branches show possible actions and consequences. Nodes (or forks) show decisions or chance points. Each branch (or possibility) commences at a node.

124 When people make decisions, they often use many of the processes that a Decision Tree “forces” them to use. These include (a) put things in writing, (b) group “like” things together, (c) write down possible advantages and disadvantages of decision (objectives) (d) record systematically the probability of each result or chance factor occurring –

125 The above processes provide the following advantages: Planners can (a) reduce chance of forgetting, (b) find more possibilities, (c) avoid choosingbetween possibilities until they give more time to “find possibilities”, (d) perceive interrelationships and patterns more easily, (e) discuss the situation with other people more-easily, (f) evaluate better the results and consequences because of the probabilities recorded.

126 When people make important decisions they do many things which Decision Trees help them to do. Thus, they help ensure that people use many of the good approaches to decision making.

127 However people still have difficulties in drawing Decision Trees and they avoid using them. This behaviour occurs because the discipline of forcing detailed consideration of the many complexities of a problem does not necessarily prove easy or enjoyable. It requires hard thinking work and, not unexpectedly, many people avoid this task. Besides, the delay in making a decision means the tension of “indecision” continues and people often find the temptation to get it all over and finished too great. Lack of experience in drawing Trees also leads people to avoid using them.

128 To draw a Decision Tree Planners should clearly identify the problem or decision under consideration (preferably in writing). Further, they should distinguish between (a) find possibilities and (b) evaluate them.

 

 

 

DRAWING DECISION TREES

129 The following series of steps should help to find branches for Trees: (a) Step 1: List possibilities in writing. (b) Step 2: Look for, and list (i) any common element found in two or more of the possibilities and/or (ii) any important elements that rate as part of one of the possibilities (or common elements) listed. Consider the “correctness” of the major words used for the purpose of the Tree. (c) Step 3: Look for, and record, “logical opposites” to possibilities, part-possibilities, or common elements. (d) Step 4: Re-write the possibilities in some type of Tree form and put like things together. (e) Step 5: Take each node and consider the branches recorded. Try to define all the Branches that will describe completely all possibilities from the particular node or fork of the branch. Re-draw the Tree, if necessary. (f) Step 6: Use a branch called “other” where (i) the planning has not found all possibilities from the one node and (ii) further time on this objective at that particular stage does not seem worthwhile; (g) Step 7: Consider the start of the Decision Tree. Try “move to the left”, i.e. find the “logical opposites” to the first branch.

130 A detailed example illustrates the steps. This example shows such aspects as: (a) how to use “logical opposites”, (b) some of the stages in the construction of large and complex Trees, (c) the trap of drawing inconsistent Trees, and (e) the need to define terms carefully (e.g. what people to include in the term “Buyer”)

131 Planners need to “go up the hierarchy” to check whether they have chosen the right problem.

132 The notes also discuss the following general points: (a) check that the Tree does not mix up two or more trees, (b) do not expect to draw the final Tree with the first attempt, (c) avoid extending the Tree along one horizontal branch only, (d) group similar possibilities together to avoid too many branches, (e) select the largest concept for a branch, (f) check which type of branch (decision or chance) should follow a previous branch, (g) check that two branches from the one node do not have a sequential, rather than “logical opposite”, relationship, (h) remember that the method of implementing some decisions rates as essential to consider when assessing possible results, (i) use groupings of branches which offer significant and critical classifications (e.g. break-even), (j) consider putting results branches from the one node in order of “goodness” where an objective order exists.

133 Universal Decision Trees exist and the notes show and explain one. Such Universal Trees have a general form that Planners can adapt for a variety of problems. The notes show one possibility.

 

 

 


[1] Readers familiar with the Hierarchy of Objectives should note that this step equals “going up” the Hierarchy. This approach involves asking the question: what higher-level objective (preferably use next-higher-level objective) does the objective under consideration aim to assist? Sometimes the answer obtained suggests that drawing a Tree will prove unnecessary because the Planner should abandon the objective. This type of thinking should have taken place earlier. However, even if it did, a repeat can prove useful at this stage because the problem may now have become clearer.

 

[2] This approach really belongs to Step 3 but Planners should not use the seven-step procedure too rigidly.

 

[3] These two areas really need a different classification to link with “Never Bought Product ‘X’”. The following shows one possibility - Considered: (a) No Brands, (b) One Brand Only, (c) More than One Brand.

 

Introduction

 

1 These notes aim to help people draw Decision Trees. They describe a Decision Tree and illustrate it. The notes also list what many people do when they make decisions and point out that a Decision Tree uses most of the ideas in a written and systematic way.

2 The notes discuss a detailed step-by-step approach to drawing Decision Trees and discuss a number of general points that should help the Reader to “grow better Trees”.

 

 

Decision Making & Objective Setting

 

3 These notes define decision making as – a conscious thinking process aimed at choosing between two or more things. A particular class of decision making involves objective setting: a conscious thinking process aimed at choosing between two or more objectives.

4 Planning involves two stages: (a) find objectives and (b) choose between objectives, and objective setting fits into the second stage of planning.

5 Decision Trees aim not only to help make decisions (choose objectives) but also to help find possibilities. Thus, it provides a planning technique that aims to help both stages

 

 A Description of a Decision Tree

6 A Decision Tree provides a method of depicting possible decisions and their possible outcomes in a logical arrangement and sequence. The following diagram shows one way of drawing a Decision Tree –

 

 

 

 

 

 

7 Decision Trees aim to group like factors together and to display possibilities in a logical and written arrangement. Usually they start with different branches (one for each possible decision). Each of these decision branches ends in a node. From each node come branches that show possible outcomes/results (actions by others) and/or consequences.

8 The outcomes of decisions include competitive moves as well as chance factors.

9 Example. A decision to market an improved product at a higher price may lead one or more opposition companies to decrease the price for their unimproved product.

10 A Tree aims to show “every” possibility at each decision point. However, this approach could mean a very large number of branches from some points. To avoid this situation occurring, people can group like possibilities together, until the number becomes manageable.

11 Planners can label one branch as “other”. In general, Planners should avoid this approach since planners use a Decision Tree to help search for additional answers to a problem. The use of “other” makes it simple to stop searching.

12 While some articles use the sloping branches approach used in the above Tree, the “square” approach used in later diagrams allows people to fit more detail into a given area.
What people normally do when they make decisions as compared with the drawing of a decision tree

13 People make many decisions within their own mind without ever writing things down or trying to draw any type of structure such as a Decision Tree. A consideration of the various things people do will help review the desirability, or otherwise, of using Decision Trees.

14 People often do not separate clearly the two major stages of planning (find objectives and choose between them). Sometimes this approach leads people to choose between possibilities before they give adequate time to the process of finding further possibilities.

15 Experienced people often use this distinction in their own thinking. Decision Trees help to make thetask of finding possibilities a separate stage. Thus, the technique helps to keep people working on – find possibilities, i.e. people avoid the trap of moving too quickly into the stage of choosing between possibilities.

16 Many people find it useful to write down possibilities. This approach means they cannot forget them temporarily and reading over the list often brings others to mind. Such a list also helps to separate possible actions from possible results.

17 Some classification of the possibilities in the list by grouping things that have some common elements will often help to find further possibilities. It may also help to evaluate some of the possibilities.

18 Another approach aims to evaluate possibilities more systematically. One method states: select a number of the likely possibilities and write down the advantages and disadvantages of each. This approach formalises what people often do in their own mind.

19 Sometimes people gain help in finding more possibilities by seeking someone else’s opinion. This approach means that one person must communicate ideas to another person. To help do so and make better use of the Adviser’s time, some people find it better to write down the points they have in their mind in some systematic and logical way.

20 With most complex problems, people find difficulties keeping in their head the many factor involved and their interrelationships. It does not prove easy to take each likely possibility and think about likely results. Some method of writing the material down soon becomes essential if people want to explore systematically the possibilities.

21 People need help to think more clearly about the possibilities of various results occurring. Most people do so in their thinking when they classify various possibilities as “very likely” or “unlikely” etc. However, these terms rate as rather loose and people should use more specific quantitative terms. Instead of rating the probability of outcome “X” occurring as “very high” or “very low” or “about average” etc., people should use quantitative terms with reference to factors divided into 100 parts.

22 Thus if something has a very low probability, people can say it has one chance in 100 of occurring. In this case, the number 0.01 describes the probability. The probability number 0.50 refers to something that has a 50/50 chance of occurring. Something with a very high probability of occurring would use the figure of 0.95 or 0.99.

23 Example A probability of 0.99 might occur when one action involves sacking someone. In some cases, everybody in Management might believe that a l00 probability exists that a strike will occur as a consequence of the sacking.

24 All the above things can occur in a person’s mind but this approach has disadvantages. First, forgetting takes place. Second, often people only see patterns and relationships between factors when they write them down.

25 The above section aims to show that a Decision Tree does not rate as very different from what happens in real life – for important decisions.

 

Why do people (a) have difficulty in drawing, and (b) avoid using decision trees?

26 Often people find difficulties in drawing Decision Trees. However, they would find that if they considered the same problem in their own minds they would have just as much difficulty.

27 In actual practice, they might appear to have less difficulty in using their mind. However probably this situation happens because they exclude some of the factors and their interrelationships from consideration.

28 Decision Trees rate as a useful tool for Executives. However, they encourage people to spend moretime on thinking. However, thinking involves hard work and therefore people often try to avoid it. In addition, people get easily distracted from thinking about the one problem. Further, they always have the overall excuse: “I haven’t got time”.

29 To discipline oneself to draw a Decision Tree and look at all the possible outcomes will not prove easy. In any case, it delays the actual making of a decision and most people find it very satisfying and relieving (in the short term) that they have made a decision. It means that they do not have to worry about the problem any longer. Most people put aside the possibility that they have made a wrong decision and they could have many more worries in the future. The temptation always exists to risk that their decision will not prove wrong.

30 Further, it will prove difficult to depict some situations in the form of a Decision Tree. Sometimes some type of table or graph will provide a better arrangement of some types of information involved in the decision- making process.

31 In addition, people need some skill in drawing Decision Trees and disciplining their mind to think in a particular way. People will find that drawing their first ten or twenty Decision Trees will prove more difficult and a lot more time consuming than drawing Trees when relatively skilled.

32 Thus people need to gain experience and practice in drawing Decision Trees in order to make it likely they will use them.

33 Initial attempts to draw a Decision Tree often prove rather difficult because the selection of all the possibilities and their logical grouping will prove difficult. The following points which should help in this task.

 

Some General Points

34 Before starting a Tree, Planners should identify clearly the problem or decision under consideration and, preferably, put it in writing.

35 In the early stages Planners should distinguish clearly between (a) possibilities and (b) the evaluation of them (i.e. possible results) Thus, initially, Planners should just identify possible objectives without trying to evaluate them. Evaluation often encourages people not to record possibilities that sound “bad” or “silly”. Since the Tree approach aims to record all possibilities, such an early evaluation would reduce the chance of achieving this aim.

 

A suggested procedure for finding Branches for Decision Trees

36 The steps listed below apply to finding branches for both possible objectives and possible results. In future, these notes use the word “possibilities” to apply to both objectives and results. However initially they emphasise possible objectives. Most examples refer to objectives –

37 After the initial listing of the steps, the notes discuss each step and give an example to illustrate how to use each step.

38 The steps do not necessarily follow strictly one after the other. People can use any of the steps for thinking of objectives at different times during the “growing” of a Tree.

39 Step 1 – List possibilities in writing.

40 Step 2. Look for, and list: (a) any common element found in two or more of the possibilities and/or (b) any important elements which rate as part of one of the possibilities (or common elements) listed. Consider the “correctness” of the major words used for the purpose of the Tree.

41 Step 3 Look for, and record, “logical opposites” to either possibilities, part-possibilities, or common elements. (Later sections explain the meaning of logical opposites.)

42 Step 4. Re-write the possibilities in some type of Tree form, putting like things together.

43 Step 5 Take each node and consider the branches recorded that follow it. Try to define all the branches that will describe completely all possibilities from the particular node or fork of the branch. Re-draw the Tree, if necessary.

44 Step 6. Use a branch called “other” where (a) you have not found all possibilities from the one node and (b) spending further time at that stage does not seem worthwhile.

45 Step 7. Consider the start of the Decision Tree and try “moving to the left”, i.e. find the “logical opposites” to the first branch.[1]

46 Step 8. Repeat Steps 2 to 6 and record any new possibilities found (i.e. Step 1).

47 It proves difficult to know just how long to continue this process. It will become a matter of judgement by Planners. Probably it should depend on the (apparent) completeness of the Tree and the importance of the end objective for which the Planner wants to draw the Tree.

48 The following paragraphs give more details on the steps described above.

Step 1: List possibilities in writing

49 Planners can (and should) carry out this step at any time during the drawing of a Tree. People forget possibilities quite easily; thus if a possibility comes to mind, record it.

50 Planners do not necessarily have to put any possibility in the correct place on the Tree: Planners who cannot find the right place easily should record it on the side of the page or on a separate page.

51 After Planners use each possibility (not recorded on the Tree), they should show themselves that they have done so by drawing a line through it. Some such systematic procedure should ensure that they have used all possibilities on the Tree or they have discarded them.

52 Example. To make these steps understandable the following sections include an actual example. Each step will show some aspect of the example.

53 The example has the objective “Increase sales of Product X’”. Step 1 produces the following result:

(a) Give Sales Representatives more training.

(b) Look for more Distributors to take on the product.

(c) Put on more Sales Representatives.

(d) Improve the product.

(e) Drop the price.

(f) Improve the package.

(g) Alter the Product’s name.

(h) Find other uses for the product.

(i) Collect opinions from people who buy it and find out why they buy it.

54 Many more possibilities exist but the above provides sufficient to make the example useful.

55 Step 2: Look for, and list: (a) any common element found in two or more of the possibilities and/or (b) any important elements which rate as part of one of the possibilities (or common elements) listed. Consider the “correctness” of the major words used for the purpose of the Tree.

56 One approach to finding a common element involves looking for possibilities that include the same word.

57 Example. In the above example, two possibilities include the term Sales Representatives. Another common element exists - the first three possibilities and the last one all relate to people. The other possibilities all relate to something other than people.

58 Another common element exists between “Improve the product” and “Find other uses for it”. Both relate to how people use the product.

59 Product and Package have a common element in that they both refer to physical aspects of the products. Similarly, Name and Price both refer to intangible aspects.

60 The example proves just as fruitful in looking for important elements in the original possibilities listed and some of the added later ones.

61 Example. The common element “people” contains important part elements. For example, people can include the classes Buyers or Sellers. (The Sellers include both Sales Representatives and Distributors, listed in the possibilities.) To become more useful, Planners could use the classification Buyers of Product “X” and Sellers of Product “X”. However, people have other classifications. A number of people exist who do not buy Product “X”. This thought produces a better classification – Sellers and Non-Sellers and Buyers and Non-Buyers.

62 This idea leads to the consideration of whether people exist who could sell the product and at present do not. Similarly, do people exist who rate as Non-Buyers and could become Buyers? (This point rates as the central point of the whole problem.)

63 if Planners consider the possibility of collecting opinions from people who have bought “X”, logically they should look at the opposite[2] of this group, i.e. people who have not bought the product. Planners can easily move down the sub-classification to (a) the non-buyers who considered buying it and (h) those non- buyers who have never considered buying it. Those who have never considered buying “X” divide into those who have (a) heard of the product and (b) not heard of the product.

 

 

64 Further thoughts indicate that buyers of the product also have two sub-classes: Buyers of our brand of Product and Buyers of other brands of the Product.

65 The following diagram shows the above possibilities.

Possible Actions to Take Concerning Various Factors \

Diagram 1

Diagram 2

People

 

 

Diagram 3

 

66 At this stage, the above diagrams show three separate Trees. It seems a better plan to approach the problem this way, i.e. record what does fit together quite readily and then look for combinations.

67 Example. Planners can easily combine diagrams (I) and (2) by tacking (I) on to the top line of (2). The use of diagram (3) will mean a re-writing of the top part of Tree (2). A later diagram does so.

Step 3: Look for, and record, “logical opposites” to possibilities, part-possibilities, or common elements.

68 The words “logical opposite” mean more than the opposite of something, although this term includes ordinary opposites.

69 Examples. “Up” rates as the opposite to “down”, However if Planners seek “logical opposites” then going sideways would qualify, even though “side” does not rate as the ordinary opposite of “up”. The “logical opposite” (and opposite) of “move to the left” would say “move to the right”, but “move forward” or “move back” would rate as “logical opposites” as well. In addition, the logical opposite of “remain still” exists.

70 Finding “logical opposites” requires some imagination but with a little experience and practice the idea will prove easy to use. Further, many situations exist in which common “logical opposites” often occur. These opposites will help to draw a Tree for such situations.

 

 

71 Example. In the example, consider the possibility of - add to the number of Sales Representatives (or Distributors). The common or universal “logical opposites” rate as -
“increase”, “no change”, and “decrease” – with respect to the number of people concerned. The same “logical opposites” apply to price.

72 In considering “Find other uses for the Product”, the two-part classification comes to mind – Present use, Possible uses. Possible uses means future uses. This stage does not include the “logical opposite” – consider past uses which today’s consumer do not use. Probably this idea will not prove very useful but it could happen that a Company could try to revitalise a forgotten past use (perhaps after amendment) and gain acceptance (e.g. when should someone next try to re-introduce the Hula Hoop or the Yo-yo?)

 

Step 4: Re-write the possibilities in some type of tree form – putting like things together.

73 The Tree on the next page rates as appropriate. Readers should understand that this topic equals a very large one (and a large Tree) and it does not prove easy to draw.

74 The Tree on the next page merely sets out various possibilities in one systematic and logical manner. It does not specify what action to take, simply that someone could take some action with the various items identified. The specific aspects identified such as “increase our Distributors or our Sales Representatives or the Price” rate as the only specific actions listed. The rest merely identify areas.

75 The actual things that someone could do in the specific areas listed need consideration. The following possibility (already identified) exists: “collect opinions from people who buy the product and find out why they buy it”. The above analysis has shown that a number of other people exist from whom it may prove worthwhile to collect opinions. It may prove desirable to draw a Tree for this particular area. At all events, it will provide another example.

76 In the actual diagram above, the four branches: “Non-Sellers”, “Sellers of Product ‘X’”, “Buyers of Product ‘Y’”, and “Non-Buyers of Product ‘X’” do not rate as mutually exclusive. Someone could fit into the classes of both a non-seller of Product “X” and a buyer. A number of other pairs such as this pair exist. These notes ignore this problem; but will consider the actual problem of looking at buyers from whom to collect opinions. The first attempt at looking at the aspect more logically gives the following diagram.

 

 

 

 

 

 

 

 

Diagram 4

Take

Action

With

Res-

pect

to -

Diagram 5

77 However even the above does not rate as completely consistent because a buyer of Product “X” who has never heard of Product “X” cannot exist. Therefore, the Tree needs re-drawing – as occurs later.

78 At this stage, an example of a different approach should prove useful. Sometimes a diagram will help to clarify a person’s thinking better than a Tree. The following diagram aimed to help see what possibilities exist. It shows the relationship between (a) Buyers/Non- Buyers and (b) people who have considered the various brands of the products or have not considered one or more of them. The central part of the diagram shows whether the combination can, or cannot, occur.

Diagram 6

 

Considered Brand Bought

Considered Brand Bought and Others

Never Considered the Product

Buyers of Product “X” Brand “A”

Must occur

Possible

Impossible

Buyers of Product “X” Other Brands

Must occur

Possible

Impossible

Never bought Product “X”

Impossible[3]

Impossible3

Possible

 

79 With Diagram 6 completed it proves easier to re-draw the Tree to get a classification which does not overlap in its categories. Diagram 7 shows this tree.

Diagram 7

 

80 Even the above diagram does not allow for “Buyer of both Product “X” and other brands”.

81 A Planner could add more detail to the above diagram in the area marked (A) by using the following:

Diagram 8

 

 

 

82 A Planner could sub divide the branch “Never bought Product “X” in a different way as shown in the following diagram

 Diagram 9

83 These notes have not defined the word “Buyer” as used in the diagram above. A Planner would introduce a further complexity by talking about various levels of buying strength.

84 Example. Regular buyers over the last two years, buyers who have just started buying the product, buyers who have just changed from one brand to our brand, and so on. The possible complexities will become almost infinite. The desirability of taking them to this extreme would depend on the activities involved and the importance of the whole objective in this area in particular.

85 At this stage Readers may well ask – Just where will this approach lead? and Will it prove useful?

86 First, it does identify the people from whom investigators can seek opinions. However – could a Planner have reached the classifications in the latter diagrams and understood some of their relationships by using a better/shorter method? Would a less-complete classification prove just as useful?

87 The next stage would become – decide what opinions to collect. The following question shows the scope: “Why did you buy, or did not buy, the particular product and/or particular brand of the product?”

88 Once someone has identified the classifications then someone can aim to evaluate the possible results of the various types of questions to the various classes of people. Perhaps researchers will not need opinions in a particular market research from people who have never considered buying the product. Early in the interview once an Interviewer identifies a Respondent as belonging to this particular category, they would ask no more questions. It can also occur that certain types of questions will prove useless to ask of people who have never considered “our Brand”.

89 With some grouping of these classes of people and the addition of why they did (or did not) buy, it will prove easier to look at the possibilities of various marketing activities having different effects.

90 Example Investgators could look at the activity of altering the price from the viewpoint of these various classes of people. They could try to assess the likelihood of them buying more, less, or having little effect - in relation to a price increase, or decrease or no change.

91 The possibility “Collect opinions, etc.” has proved an extremely complex one. However, this type of situation often happens in an organisation: someone asks someone to do something that rates as very broad and unless someone works out the details often what people do proves wasted. Quite often, the actual amount of work exceeds what the original person suggesting the approach envisaged. Sometimes people carry out the broad objective quite poorly. People should not feel surprised at the useless results that often occur.

 

Step 7: Consider the start of the Decision Tree and try “moving to the left”, i.e. find the “logical opposites” to the first branch.

92 The discussion of the example has illustrated the processes of Step 5 and Step 6 but it would take a great deal more effort to draw the whole Tree for the objective of the example and it would cover a very large sheet of paper.

93 Example. It will prove simple to find some “logical opposites” of the objective - “Increase sales of Product ‘X’”, The possibilities include: (a) maintain the present sales, (b) decrease the sales, or (c) increase them. The area of decreasing sales has an end to it in the sense that the objective could state – achieve no sales. This objective equals deleting the product from the company’s operations. Although Planners may have considered this objective, they should keep it well to the fore since it will prove easy to forget or overlook this “obvious” possibility.

94 Investigators could first aim to establish the likelihood that sales of the product can increase. They can compare two objectives – (a) spend money for market research on finding out why people buy and do not buy the product versus (b) ask questions about the real need which the product has in the past satisfied and will anything change (reduce) this need and will other products satisfy it better,

General Points

95 Apart from the steps discussed above the following general points should help in drawing Decision Trees.

96 Check that you aim to draw only one Tree at a time. Sometimes a number of problems will exist at the same time and interact. A Tree may prove difficult to draw because no one has established the separating point between the problems. Planners should do so.

97 Do not try too hard to fit all the possibilities into the initial Tree. Planners who have difficulty finding a place for a possibility in a Tree should write it down on one side of the page. At a later stage, it may prove much easier to fit it into the structure – because the drawing of the Tree has established a better understanding of the whole situation. Eventually it must prove possible to fit it in or feel sure that it belongs to a different problem.

98 With complex problems, it will often help to draw “small” Trees for different parts of the problem and then try to put them together.

99 Develop the Tree from all nodes on each vertical level – rather than extending the Tree on one horizontal level much further than at other levels.

100 In drawing the possibilities from each node, group together similar possibilities. Probably no node should have more than about five or six branches coming from it.

101 Use the largest concept or idea in deciding on a branch.

102 Examples. Profitability rates as a larger concept that sales since the interaction between sales and expenses gives profit. “Go home” (for a person already at work) gives a “smaller” idea than “Leave work”. In this case “Go home” rates as the smaller concept since “Go home” means that the person must “Leave work”, but the person could also “Go to the races” or “Go out to lunch”. The larger concept “Leave work” covers all the possibilities but “Leave work” does not necessarily mean that the person must “Go home”.

103 Decide when another decision node(as opposed to a chance node) should follow an originaldecision node.Similarly decide when a chance node should follow another chance node.

104 Some two-part decisions only help to group possibilities logically. Thus some things to do must follow a “do something” (versus “do nothing”) node.

105 A variation of this point states – some possible decisions have a sequential link; they can never exist as alternatives.

106 Example. Probably the objectives - “Phone the Supervisor” and “Ignore the different instructions given” rate as sequential. The following diagram shows an appropriate Tree of the possibilities.

Diagram 10.

107 A slightly-different form might improve the tree:

Diagram 11

108 The diagram in a different form shows more clearly that the action of informing someone rates as independent of the decision to accept or reject different instructions.

In the context of - I will inform my Supervisor, I will not carry out (i.e. I reject) his instructions, the term ignore does not fit. Usually people ignore instructions but do not tell their Supervisor that they intend to ignore the instructions. The term “passive resistance” describes this situation

109 Another reason exists which supports the idea that a chance factor should not necessarily follow a decision node. Sometimes the particular methods used to implement a decision will prove almost as significant as the decision. In this case, the reactions to the decision should really take into account the means of implementing the decision.

110 This point holds true particularly in personnel problems. However with broader scale objectives (e.g. building a new factory or launch a new product) the implementation rates as important, but often it will prove useful to assume that the implementation will proceed with average efficiency. Where implementation largely depends on one or a few factors (e.g. the persuasiveness of one person) then Planners should link the methods with the implementation of the decision.

111 Look for critical and significant possibilitieswhen grouping and arranging possibilities.

112 Example. In looking at consequences of a decision, a critical point often becomes whether the result will reach a break-even point (where income equals expenditure). Thus, the three branches from a consequence or chance node might show (a) makes a profit, (b) breaks even, (c) makes a loss.

113 One interesting example occurs with the result – “effect on productivity”. It can involve increase, no change, and decrease. However, sometimes Planners should look carefully at a special point of decrease, i.e. nil productivity. In certain industrial situations, this point means that a strike has resulted.

114 Recording branches in order on some scale.

115 Examples. Record good, bad, or neutral results in order: Good, neutral, bad (or bad, neutral, good). High Profit, Medium Profit, Break Even, Loss shows another set of possibilities arranged in order

 116 Sometimes Planners should put all the good possibilities at a higher (branch) level than poor ones. This point assumes that Planners can evaluate easily the “goodness” of a branch. In some cases, it can (e.g. profit versus loss); in others it cannot. (This suggestion also tends to oppose the previous suggestion of not evaluating possibilities but sometimes the evaluation rates as obvious and not a matter of personal opinion.)

117 Remember that Planners draw a Tree to help make decisions. The provision of many branches or their elaboration can waste time and money. However once someone makes particular decisions then it will prove appropriate to elaborate (draw another Tree) on the particular branches chosen.

 

Universal Trees

118 Some Trees exist which rate as appropriate for a number of problems. Their structure may need adaptation for particular problems; they will certainly need translation.

 

 

 

 

 

 

 

 

119 The following shows one example.

“Universal” Decision Tree Form for Decisions

120 This Tree could not show easily all the possible consequences that might arise from any one decision since a large number of decisions exist. However, Planners could select the four or six “best” and draw the consequences of these decisions.

121 For Planners who select the decision “collect further information”, methods of continuing or abandoning the objective will not prove appropriate until someone obtains the “further information”. One method of continuing the Tree would involve listing the likely broad results that might come from the further investigation. Then it could show decisions on how to continue or abandon. However probably this approach would prove inappropriate unless the whole matter rated as so urgent that Planners should spend time in planning to enable a quick start on continuing the problem immediately they obtained “further information”. Unless this situation existed, it would prove better to defer further thinking until someone obtained the “further information”.

 

Summary

122 A Decision Tree shows possible decisions and their possible outcomes written out in a logical arrangement and sequence. It aims to help both stages of planning, i.e. (a) find possibilities (objectives) and (b) choose between possibilities. These notes call the second stage of planning “decision making”.

123 In Decision Trees, Branches show possible actions and consequences. Nodes (or forks) show decisions or chance points. Each branch (or possibility) commences at a node.

124 When people make decisions, they often use many of the processes that a Decision Tree “forces” them to use. These include (a) put things in writing, (b) group “like” things together, (c) write down possible advantages and disadvantages of decision (objectives) (d) record systematically the probability of each result or chance factor occurring –

125 The above processes provide the following advantages: Planners can (a) reduce chance of forgetting, (b) find more possibilities, (c) avoid choosingbetween possibilities until they give more time to “find possibilities”, (d) perceive interrelationships and patterns more easily, (e) discuss the situation with other people more-easily, (f) evaluate better the results and consequences because of the probabilities recorded.

126 When people make important decisions they do many things which Decision Trees help them to do. Thus, they help ensure that people use many of the good approaches to decision making.

127 However people still have difficulties in drawing Decision Trees and they avoid using them. This behaviour occurs because the discipline of forcing detailed consideration of the many complexities of a problem does not necessarily prove easy or enjoyable. It requires hard thinking work and, not unexpectedly, many people avoid this task. Besides, the delay in making a decision means the tension of “indecision” continues and people often find the temptation to get it all over and finished too great. Lack of experience in drawing Trees also leads people to avoid using them.

128 To draw a Decision Tree Planners should clearly identify the problem or decision under consideration (preferably in writing). Further, they should distinguish between (a) find possibilities and (b) evaluate them.

 

 

 

DRAWING DECISION TREES

129 The following series of steps should help to find branches for Trees: (a) Step 1: List possibilities in writing. (b) Step 2: Look for, and list (i) any common element found in two or more of the possibilities and/or (ii) any important elements that rate as part of one of the possibilities (or common elements) listed. Consider the “correctness” of the major words used for the purpose of the Tree. (c) Step 3: Look for, and record, “logical opposites” to possibilities, part-possibilities, or common elements. (d) Step 4: Re-write the possibilities in some type of Tree form and put like things together. (e) Step 5: Take each node and consider the branches recorded. Try to define all the Branches that will describe completely all possibilities from the particular node or fork of the branch. Re-draw the Tree, if necessary. (f) Step 6: Use a branch called “other” where (i) the planning has not found all possibilities from the one node and (ii) further time on this objective at that particular stage does not seem worthwhile; (g) Step 7: Consider the start of the Decision Tree. Try “move to the left”, i.e. find the “logical opposites” to the first branch.

130 A detailed example illustrates the steps. This example shows such aspects as: (a) how to use “logical opposites”, (b) some of the stages in the construction of large and complex Trees, (c) the trap of drawing inconsistent Trees, and (e) the need to define terms carefully (e.g. what people to include in the term “Buyer”)

131 Planners need to “go up the hierarchy” to check whether they have chosen the right problem.

132 The notes also discuss the following general points: (a) check that the Tree does not mix up two or more trees, (b) do not expect to draw the final Tree with the first attempt, (c) avoid extending the Tree along one horizontal branch only, (d) group similar possibilities together to avoid too many branches, (e) select the largest concept for a branch, (f) check which type of branch (decision or chance) should follow a previous branch, (g) check that two branches from the one node do not have a sequential, rather than “logical opposite”, relationship, (h) remember that the method of implementing some decisions rates as essential to consider when assessing possible results, (i) use groupings of branches which offer significant and critical classifications (e.g. break-even), (j) consider putting results branches from the one node in order of “goodness” where an objective order exists.

133 Universal Decision Trees exist and the notes show and explain one. Such Universal Trees have a general form that Planners can adapt for a variety of problems. The notes show one possibility.

 

 

 


[1] Readers familiar with the Hierarchy of Objectives should note that this step equals “going up” the Hierarchy. This approach involves asking the question: what higher-level objective (preferably use next-higher-level objective) does the objective under consideration aim to assist? Sometimes the answer obtained suggests that drawing a Tree will prove unnecessary because the Planner should abandon the objective. This type of thinking should have taken place earlier. However, even if it did, a repeat can prove useful at this stage because the problem may now have become clearer.

 

[2] This approach really belongs to Step 3 but Planners should not use the seven-step procedure too rigidly.

 

[3] These two areas really need a different classification to link with “Never Bought Product ‘X’”. The following shows one possibility - Considered: (a) No Brands, (b) One Brand Only, (c) More than One Brand.

 

Exercise – Expected Values for Decision Tree – Exercise 10

  

Table 1  – EXPECTED VALUE OF DECISION TWO

          EXPECTED  

DECISION

DEMAND

p

PROFIT

p x PROFIT    

ADDITION

-

COST

=

VALUE  

High

0.8

4

3.2

Expand

3.6

-

2.5

=

1.1

Low

0.2

2

0.4

No

High

0.8

3

2.4

Explanation

2.9

-

Nil

=

2.9

Low

0.2

2.5

0.5

The preferred decision for Decision 2 becomes “No Expansion”.

TABLE 2  – EXPECTED VALUE OF DECISION ONE

                                      EXPECTED

DECISION

DEMAND

p

PROFIT

p x PROFIT

ADDITION

-

COST VALUE  

High

0.6

9

5.4

Large

High Low

0.1

3

0.3

6.00

-

3.5

=

2.5

Plant

Low

0.3

1

0.3

High

0.7

5.9

4.13

Small

5.27

-

2.0

=

3.27

Plant

Low

0.3

3.8

1.14

The preferred decision of Decision 2 becomes $2.9 million for a three-year period plus $3 million for the first two years.

 

 

 

TABLE 1  – EXPECTED VALUE OF DECISION 2
          EXPECTED  

DECISION

DEMAND

p

PROFIT p x PROFIT     ADDITION

ADDITION

-

COST

=

VALUE  

High

0.8

4

3.2

Expand

3.6

-

2.5

=

1.1

Low

0.2

2

0.4

No

High

0.8

3

2.4

Explanation

2.9

-

Nil

=

2.9

Low

0.2

2.5

0.5

The preferred decision for Decision 2 becomes “No Expansion”.
TABLE 2  – EXPECTED VALUE OF DECISION 1
                                      EXPECTED

DECISION

DEMAND

p

PROFIT

px PROFIT

ADDITION

-

COST VALUE  

High

0.6

9

5.4

Large

High Low

0.1

3

0.3

6.00

-

3.5

=

2.5

Plant

Low

0.3

1

0.3

High

0.7

5.9

4.13

Small

5.27

-

2.0

=

3.27

Plant

Low

0.3

3.8

1.14

The preferred decision of Decision 2 becomes $2.9 million for a thre-year period plus
      $3 million for the first two years.

 

Drawing Decision Trees 1

Introduction

 

1 These notes aim to help people draw Decision Trees. They describe a Decision Tree and illustrate it. The notes also list what many people do when they make decisions and point out that a Decision Tree uses most of the ideas in a written and systematic way.

2 The notes discuss a detailed step-by-step approach to drawing Decision Trees and discuss a number of general points that should help the Reader to “grow better Trees”.

 

 

Decision Making & Objective Setting

 

3 These notes define decision making as – a conscious thinking process aimed at choosing between two or more things. A particular class of decision making involves objective setting: a conscious thinking process aimed at choosing between two or more objectives.

4 Planning involves two stages: (a) find objectives and (b) choose between objectives, and objective setting fits into the second stage of planning.

5 Decision Trees aim not only to help make decisions (choose objectives) but also to help find possibilities. Thus, it provides a planning technique that aims to help both stages

 

 A Description of a Decision Tree

6 A Decision Tree provides a method of depicting possible decisions and their possible outcomes in a logical arrangement and sequence. The following diagram shows one way of drawing a Decision Tree –

 

 

7 Decision Trees aim to group like factors together and to display possibilities in a logical and written arrangement. Usually they start with different branches (one for each possible decision). Each of these decision branches ends in a node. From each node come branches that show possible outcomes/results (actions by others) and/or consequences.

 

8 The outcomes of decisions include competitive moves as well as chance factors.

9 Example. A decision to market an improved product at a higher price may lead one or more opposition companies to decrease the price for their unimproved product.

10 A Tree aims to show “every” possibility at each decision point. However, this approach could mean a very large number of branches from some points. To avoid this situation occurring, people can group like possibilities together, until the number becomes manageable.

11 Planners can label one branch as “other”. In general, Planners should avoid this approach since planners use a Decision Tree to help search for additional answers to a problem. The use of “other” makes it simple to stop searching.

12 While some articles use the sloping branches approach used in the above Tree, the “square” approach used in later diagrams allows people to fit more detail into a given area.
What people normally do when they make decisions as compared with the drawing of a decision tree

13 People make many decisions within their own mind without ever writing things down or trying to draw any type of structure such as a Decision Tree. A consideration of the various things people do will help review the desirability, or otherwise, of using Decision Trees.

14 People often do not separate clearly the two major stages of planning (find objectives and choose between them). Sometimes this approach leads people to choose between possibilities before they give adequate time to the process of finding further possibilities.

15 Experienced people often use this distinction in their own thinking. Decision Trees help to make thetask of finding possibilities a separate stage. Thus, the technique helps to keep people working on – find possibilities, i.e. people avoid the trap of moving too quickly into the stage of choosing between possibilities.

16 Many people find it useful to write down possibilities. This approach means they cannot forget them temporarily and reading over the list often brings others to mind. Such a list also helps to separate possible actions from possible results.

17 Some classification of the possibilities in the list by grouping things that have some common elements will often help to find further possibilities. It may also help to evaluate some of the possibilities.

18 Another approach aims to evaluate possibilities more systematically. One method states: select a number of the likely possibilities and write down the advantages and disadvantages of each. This approach formalises what people often do in their own mind.

19 Sometimes people gain help in finding more possibilities by seeking someone else’s opinion. This approach means that one person must communicate ideas to another person. To help do so and make better use of the Adviser’s time, some people find it better to write down the points they have in their mind in some systematic and logical way.

20 With most complex problems, people find difficulties keeping in their head the many factor involved and their interrelationships. It does not prove easy to take each likely possibility and think about likely results. Some method of writing the material down soon becomes essential if people want to explore systematically the possibilities.

21 People need help to think more clearly about the possibilities of various results occurring. Most people do so in their thinking when they classify various possibilities as “very likely” or “unlikely” etc. However, these terms rate as rather loose and people should use more specific quantitative terms. Instead of rating the probability of outcome “X” occurring as “very high” or “very low” or “about average” etc., people should use quantitative terms with reference to factors divided into 100 parts.

22 Thus if something has a very low probability, people can say it has one chance in 100 of occurring. In this case, the number 0.01 describes the probability. The probability number 0.50 refers to something that has a 50/50 chance of occurring. Something with a very high probability of occurring would use the figure of 0.95 or 0.99.

23 Example A probability of 0.99 might occur when one action involves sacking someone. In some cases, everybody in Management might believe that a l00 probability exists that a strike will occur as a consequence of the sacking.

24 All the above things can occur in a person’s mind but this approach has disadvantages. First, forgetting takes place. Second, often people only see patterns and relationships between factors when they write them down.

25 The above section aims to show that a Decision Tree does not rate as very different from what happens in real life – for important decisions.

 

Why do people (a) have difficulty in drawing, and (b) avoid using decision trees?

26 Often people find difficulties in drawing Decision Trees. However, they would find that if they considered the same problem in their own minds they would have just as much difficulty.

27 In actual practice, they might appear to have less difficulty in using their mind. However probably this situation happens because they exclude some of the factors and their interrelationships from consideration.

28 Decision Trees rate as a useful tool for Executives. However, they encourage people to spend moretime on thinking. However, thinking involves hard work and therefore people often try to avoid it. In addition, people get easily distracted from thinking about the one problem. Further, they always have the overall excuse: “I haven’t got time”.

29 To discipline oneself to draw a Decision Tree and look at all the possible outcomes will not prove easy. In any case, it delays the actual making of a decision and most people find it very satisfying and relieving (in the short term) that they have made a decision. It means that they do not have to worry about the problem any longer. Most people put aside the possibility that they have made a wrong decision and they could have many more worries in the future. The temptation always exists to risk that their decision will not prove wrong.

30 Further, it will prove difficult to depict some situations in the form of a Decision Tree. Sometimes some type of table or graph will provide a better arrangement of some types of information involved in the decision- making process.

31 In addition, people need some skill in drawing Decision Trees and disciplining their mind to think in a particular way. People will find that drawing their first ten or twenty Decision Trees will prove more difficult and a lot more time consuming than drawing Trees when relatively skilled.

32 Thus people need to gain experience and practice in drawing Decision Trees in order to make it likely they will use them.

33 Initial attempts to draw a Decision Tree often prove rather difficult because the selection of all the possibilities and their logical grouping will prove difficult. The following points which should help in this task.

 

Some General Points

34 Before starting a Tree, Planners should identify clearly the problem or decision under consideration and, preferably, put it in writing.

35 In the early stages Planners should distinguish clearly between (a) possibilities and (b) the evaluation of them (i.e. possible results) Thus, initially, Planners should just identify possible objectives without trying to evaluate them. Evaluation often encourages people not to record possibilities that sound “bad” or “silly”. Since the Tree approach aims to record all possibilities, such an early evaluation would reduce the chance of achieving this aim.

 

A suggested procedure for finding Branches for Decision Trees

36 The steps listed below apply to finding branches for both possible objectives and possible results. In future, these notes use the word “possibilities” to apply to both objectives and results. However initially they emphasise possible objectives. Most examples refer to objectives –

37 After the initial listing of the steps, the notes discuss each step and give an example to illustrate how to use each step.

38 The steps do not necessarily follow strictly one after the other. People can use any of the steps for thinking of objectives at different times during the “growing” of a Tree.

39 Step 1 – List possibilities in writing.

40 Step 2. Look for, and list: (a) any common element found in two or more of the possibilities and/or (b) any important elements which rate as part of one of the possibilities (or common elements) listed. Consider the “correctness” of the major words used for the purpose of the Tree.

41 Step 3 Look for, and record, “logical opposites” to either possibilities, part-possibilities, or common elements. (Later sections explain the meaning of logical opposites.)

42 Step 4. Re-write the possibilities in some type of Tree form, putting like things together.

43 Step 5 Take each node and consider the branches recorded that follow it. Try to define all the branches that will describe completely all possibilities from the particular node or fork of the branch. Re-draw the Tree, if necessary.

44 Step 6. Use a branch called “other” where (a) you have not found all possibilities from the one node and (b) spending further time at that stage does not seem worthwhile.

45 Step 7. Consider the start of the Decision Tree and try “moving to the left”, i.e. find the “logical opposites” to the first branch.[1]

46 Step 8. Repeat Steps 2 to 6 and record any new possibilities found (i.e. Step 1).

47 It proves difficult to know just how long to continue this process. It will become a matter of judgement by Planners. Probably it should depend on the (apparent) completeness of the Tree and the importance of the end objective for which the Planner wants to draw the Tree.

48 The following paragraphs give more details on the steps described above.

Step 1: List possibilities in writing

49 Planners can (and should) carry out this step at any time during the drawing of a Tree. People forget possibilities quite easily; thus if a possibility comes to mind, record it.

50 Planners do not necessarily have to put any possibility in the correct place on the Tree: Planners who cannot find the right place easily should record it on the side of the page or on a separate page.

51 After Planners use each possibility (not recorded on the Tree), they should show themselves that they have done so by drawing a line through it. Some such systematic procedure should ensure that they have used all possibilities on the Tree or they have discarded them.

52 Example. To make these steps understandable the following sections include an actual example. Each step will show some aspect of the example.

53 The example has the objective “Increase sales of Product X’”. Step 1 produces the following result:

(a) Give Sales Representatives more training.

(b) Look for more Distributors to take on the product.

(c) Put on more Sales Representatives.

(d) Improve the product.

(e) Drop the price.

(f) Improve the package.

(g) Alter the Product’s name.

(h) Find other uses for the product.

(i) Collect opinions from people who buy it and find out why they buy it.

54 Many more possibilities exist but the above provides sufficient to make the example useful.

55 Step 2: Look for, and list: (a) any common element found in two or more of the possibilities and/or (b) any important elements which rate as part of one of the possibilities (or common elements) listed. Consider the “correctness” of the major words used for the purpose of the Tree.

56 One approach to finding a common element involves looking for possibilities that include the same word.

57 Example. In the above example, two possibilities include the term Sales Representatives. Another common element exists - the first three possibilities and the last one all relate to people. The other possibilities all relate to something other than people.

58 Another common element exists between “Improve the product” and “Find other uses for it”. Both relate to how people use the product.

59 Product and Package have a common element in that they both refer to physical aspects of the products. Similarly, Name and Price both refer to intangible aspects.

60 The example proves just as fruitful in looking for important elements in the original possibilities listed and some of the added later ones.

61 Example. The common element “people” contains important part elements. For example, people can include the classes Buyers or Sellers. (The Sellers include both Sales Representatives and Distributors, listed in the possibilities.) To become more useful, Planners could use the classification Buyers of Product “X” and Sellers of Product “X”. However, people have other classifications. A number of people exist who do not buy Product “X”. This thought produces a better classification – Sellers and Non-Sellers and Buyers and Non-Buyers.

62 This idea leads to the consideration of whether people exist who could sell the product and at present do not. Similarly, do people exist who rate as Non-Buyers and could become Buyers? (This point rates as the central point of the whole problem.)

63 if Planners consider the possibility of collecting opinions from people who have bought “X”, logically they should look at the opposite[2] of this group, i.e. people who have not bought the product. Planners can easily move down the sub-classification to (a) the non-buyers who considered buying it and (h) those non- buyers who have never considered buying it. Those who have never considered buying “X” divide into those who have (a) heard of the product and (b) not heard of the product.

 

 

64 Further thoughts indicate that buyers of the product also have two sub-classes: Buyers of our brand of Product and Buyers of other brands of the Product.

65 The following diagram shows the above possibilities.

Possible Actions to Take Concerning Various Factors \

Diagram 1

Diagram 2

People

 

 

Diagram 3

 

66 At this stage, the above diagrams show three separate Trees. It seems a better plan to approach the problem this way, i.e. record what does fit together quite readily and then look for combinations.

67 Example. Planners can easily combine diagrams (I) and (2) by tacking (I) on to the top line of (2). The use of diagram (3) will mean a re-writing of the top part of Tree (2). A later diagram does so.

Step 3: Look for, and record, “logical opposites” to possibilities, part-possibilities, or common elements.

68 The words “logical opposite” mean more than the opposite of something, although this term includes ordinary opposites.

69 Examples. “Up” rates as the opposite to “down”, However if Planners seek “logical opposites” then going sideways would qualify, even though “side” does not rate as the ordinary opposite of “up”. The “logical opposite” (and opposite) of “move to the left” would say “move to the right”, but “move forward” or “move back” would rate as “logical opposites” as well. In addition, the logical opposite of “remain still” exists.

70 Finding “logical opposites” requires some imagination but with a little experience and practice the idea will prove easy to use. Further, many situations exist in which common “logical opposites” often occur. These opposites will help to draw a Tree for such situations.

 

 

71 Example. In the example, consider the possibility of - add to the number of Sales Representatives (or Distributors). The common or universal “logical opposites” rate as -
“increase”, “no change”, and “decrease” – with respect to the number of people concerned. The same “logical opposites” apply to price.

72 In considering “Find other uses for the Product”, the two-part classification comes to mind – Present use, Possible uses. Possible uses means future uses. This stage does not include the “logical opposite” – consider past uses which today’s consumer do not use. Probably this idea will not prove very useful but it could happen that a Company could try to revitalise a forgotten past use (perhaps after amendment) and gain acceptance (e.g. when should someone next try to re-introduce the Hula Hoop or the Yo-yo?)

 

Step 4: Re-write the possibilities in some type of tree form – putting like things together.

73 The Tree on the next page rates as appropriate. Readers should understand that this topic equals a very large one (and a large Tree) and it does not prove easy to draw.

74 The Tree on the next page merely sets out various possibilities in one systematic and logical manner. It does not specify what action to take, simply that someone could take some action with the various items identified. The specific aspects identified such as “increase our Distributors or our Sales Representatives or the Price” rate as the only specific actions listed. The rest merely identify areas.

75 The actual things that someone could do in the specific areas listed need consideration. The following possibility (already identified) exists: “collect opinions from people who buy the product and find out why they buy it”. The above analysis has shown that a number of other people exist from whom it may prove worthwhile to collect opinions. It may prove desirable to draw a Tree for this particular area. At all events, it will provide another example.

76 In the actual diagram above, the four branches: “Non-Sellers”, “Sellers of Product ‘X’”, “Buyers of Product ‘Y’”, and “Non-Buyers of Product ‘X’” do not rate as mutually exclusive. Someone could fit into the classes of both a non-seller of Product “X” and a buyer. A number of other pairs such as this pair exist. These notes ignore this problem; but will consider the actual problem of looking at buyers from whom to collect opinions. The first attempt at looking at the aspect more logically gives the following diagram.

 

 

 

 

 

 

 

 

Diagram 4

Take

Action

With

Res-

pect

to -

Diagram 5

77 However even the above does not rate as completely consistent because a buyer of Product “X” who has never heard of Product “X” cannot exist. Therefore, the Tree needs re-drawing – as occurs later.

78 At this stage, an example of a different approach should prove useful. Sometimes a diagram will help to clarify a person’s thinking better than a Tree. The following diagram aimed to help see what possibilities exist. It shows the relationship between (a) Buyers/Non- Buyers and (b) people who have considered the various brands of the products or have not considered one or more of them. The central part of the diagram shows whether the combination can, or cannot, occur.

Diagram 6

 

Considered Brand Bought

Considered Brand Bought and Others

Never Considered the Product

Buyers of Product “X” Brand “A”

Must occur

Possible

Impossible

Buyers of Product “X” Other Brands

Must occur

Possible

Impossible

Never bought Product “X”

Impossible[3]

Impossible3

Possible

 

79 With Diagram 6 completed it proves easier to re-draw the Tree to get a classification which does not overlap in its categories. Diagram 7 shows this tree.

Diagram 7

 

80 Even the above diagram does not allow for “Buyer of both Product “X” and other brands”.

81 A Planner could add more detail to the above diagram in the area marked (A) by using the following:

Diagram 8

 

 

 

82 A Planner could sub divide the branch “Never bought Product “X” in a different way as shown in the following diagram

 Diagram 9

83 These notes have not defined the word “Buyer” as used in the diagram above. A Planner would introduce a further complexity by talking about various levels of buying strength.

84 Example. Regular buyers over the last two years, buyers who have just started buying the product, buyers who have just changed from one brand to our brand, and so on. The possible complexities will become almost infinite. The desirability of taking them to this extreme would depend on the activities involved and the importance of the whole objective in this area in particular.

85 At this stage Readers may well ask – Just where will this approach lead? and Will it prove useful?

86 First, it does identify the people from whom investigators can seek opinions. However – could a Planner have reached the classifications in the latter diagrams and understood some of their relationships by using a better/shorter method? Would a less-complete classification prove just as useful?

87 The next stage would become – decide what opinions to collect. The following question shows the scope: “Why did you buy, or did not buy, the particular product and/or particular brand of the product?”

88 Once someone has identified the classifications then someone can aim to evaluate the possible results of the various types of questions to the various classes of people. Perhaps researchers will not need opinions in a particular market research from people who have never considered buying the product. Early in the interview once an Interviewer identifies a Respondent as belonging to this particular category, they would ask no more questions. It can also occur that certain types of questions will prove useless to ask of people who have never considered “our Brand”.

89 With some grouping of these classes of people and the addition of why they did (or did not) buy, it will prove easier to look at the possibilities of various marketing activities having different effects.

90 Example Investgators could look at the activity of altering the price from the viewpoint of these various classes of people. They could try to assess the likelihood of them buying more, less, or having little effect - in relation to a price increase, or decrease or no change.

91 The possibility “Collect opinions, etc.” has proved an extremely complex one. However, this type of situation often happens in an organisation: someone asks someone to do something that rates as very broad and unless someone works out the details often what people do proves wasted. Quite often, the actual amount of work exceeds what the original person suggesting the approach envisaged. Sometimes people carry out the broad objective quite poorly. People should not feel surprised at the useless results that often occur.

 

Step 7: Consider the start of the Decision Tree and try “moving to the left”, i.e. find the “logical opposites” to the first branch.

92 The discussion of the example has illustrated the processes of Step 5 and Step 6 but it would take a great deal more effort to draw the whole Tree for the objective of the example and it would cover a very large sheet of paper.

93 Example. It will prove simple to find some “logical opposites” of the objective - “Increase sales of Product ‘X’”, The possibilities include: (a) maintain the present sales, (b) decrease the sales, or (c) increase them. The area of decreasing sales has an end to it in the sense that the objective could state – achieve no sales. This objective equals deleting the product from the company’s operations. Although Planners may have considered this objective, they should keep it well to the fore since it will prove easy to forget or overlook this “obvious” possibility.

94 Investigators could first aim to establish the likelihood that sales of the product can increase. They can compare two objectives – (a) spend money for market research on finding out why people buy and do not buy the product versus (b) ask questions about the real need which the product has in the past satisfied and will anything change (reduce) this need and will other products satisfy it better,

General Points

95 Apart from the steps discussed above the following general points should help in drawing Decision Trees.

96 Check that you aim to draw only one Tree at a time. Sometimes a number of problems will exist at the same time and interact. A Tree may prove difficult to draw because no one has established the separating point between the problems. Planners should do so.

97 Do not try too hard to fit all the possibilities into the initial Tree. Planners who have difficulty finding a place for a possibility in a Tree should write it down on one side of the page. At a later stage, it may prove much easier to fit it into the structure – because the drawing of the Tree has established a better understanding of the whole situation. Eventually it must prove possible to fit it in or feel sure that it belongs to a different problem.

98 With complex problems, it will often help to draw “small” Trees for different parts of the problem and then try to put them together.

99 Develop the Tree from all nodes on each vertical level – rather than extending the Tree on one horizontal level much further than at other levels.

100 In drawing the possibilities from each node, group together similar possibilities. Probably no node should have more than about five or six branches coming from it.

101 Use the largest concept or idea in deciding on a branch.

102 Examples. Profitability rates as a larger concept that sales since the interaction between sales and expenses gives profit. “Go home” (for a person already at work) gives a “smaller” idea than “Leave work”. In this case “Go home” rates as the smaller concept since “Go home” means that the person must “Leave work”, but the person could also “Go to the races” or “Go out to lunch”. The larger concept “Leave work” covers all the possibilities but “Leave work” does not necessarily mean that the person must “Go home”.

103 Decide when another decision node(as opposed to a chance node) should follow an originaldecision node.Similarly decide when a chance node should follow another chance node.

104 Some two-part decisions only help to group possibilities logically. Thus some things to do must follow a “do something” (versus “do nothing”) node.

105 A variation of this point states – some possible decisions have a sequential link; they can never exist as alternatives.

106 Example. Probably the objectives - “Phone the Supervisor” and “Ignore the different instructions given” rate as sequential. The following diagram shows an appropriate Tree of the possibilities.

Diagram 10.

107 A slightly-different form might improve the tree:

Diagram 11

108 The diagram in a different form shows more clearly that the action of informing someone rates as independent of the decision to accept or reject different instructions.

In the context of - I will inform my Supervisor, I will not carry out (i.e. I reject) his instructions, the term ignore does not fit. Usually people ignore instructions but do not tell their Supervisor that they intend to ignore the instructions. The term “passive resistance” describes this situation

109 Another reason exists which supports the idea that a chance factor should not necessarily follow a decision node. Sometimes the particular methods used to implement a decision will prove almost as significant as the decision. In this case, the reactions to the decision should really take into account the means of implementing the decision.

110 This point holds true particularly in personnel problems. However with broader scale objectives (e.g. building a new factory or launch a new product) the implementation rates as important, but often it will prove useful to assume that the implementation will proceed with average efficiency. Where implementation largely depends on one or a few factors (e.g. the persuasiveness of one person) then Planners should link the methods with the implementation of the decision.

111 Look for critical and significant possibilitieswhen grouping and arranging possibilities.

112 Example. In looking at consequences of a decision, a critical point often becomes whether the result will reach a break-even point (where income equals expenditure). Thus, the three branches from a consequence or chance node might show (a) makes a profit, (b) breaks even, (c) makes a loss.

113 One interesting example occurs with the result – “effect on productivity”. It can involve increase, no change, and decrease. However, sometimes Planners should look carefully at a special point of decrease, i.e. nil productivity. In certain industrial situations, this point means that a strike has resulted.

114 Recording branches in order on some scale.

115 Examples. Record good, bad, or neutral results in order: Good, neutral, bad (or bad, neutral, good). High Profit, Medium Profit, Break Even, Loss shows another set of possibilities arranged in order

 116 Sometimes Planners should put all the good possibilities at a higher (branch) level than poor ones. This point assumes that Planners can evaluate easily the “goodness” of a branch. In some cases, it can (e.g. profit versus loss); in others it cannot. (This suggestion also tends to oppose the previous suggestion of not evaluating possibilities but sometimes the evaluation rates as obvious and not a matter of personal opinion.)

117 Remember that Planners draw a Tree to help make decisions. The provision of many branches or their elaboration can waste time and money. However once someone makes particular decisions then it will prove appropriate to elaborate (draw another Tree) on the particular branches chosen.

 

Universal Trees

118 Some Trees exist which rate as appropriate for a number of problems. Their structure may need adaptation for particular problems; they will certainly need translation.

 

 

 

 

 

 

 

 

119 The following shows one example.

“Universal” Decision Tree Form for Decisions

120 This Tree could not show easily all the possible consequences that might arise from any one decision since a large number of decisions exist. However, Planners could select the four or six “best” and draw the consequences of these decisions.

121 For Planners who select the decision “collect further information”, methods of continuing or abandoning the objective will not prove appropriate until someone obtains the “further information”. One method of continuing the Tree would involve listing the likely broad results that might come from the further investigation. Then it could show decisions on how to continue or abandon. However probably this approach would prove inappropriate unless the whole matter rated as so urgent that Planners should spend time in planning to enable a quick start on continuing the problem immediately they obtained “further information”. Unless this situation existed, it would prove better to defer further thinking until someone obtained the “further information”.

 

Summary

122 A Decision Tree shows possible decisions and their possible outcomes written out in a logical arrangement and sequence. It aims to help both stages of planning, i.e. (a) find possibilities (objectives) and (b) choose between possibilities. These notes call the second stage of planning “decision making”.

123 In Decision Trees, Branches show possible actions and consequences. Nodes (or forks) show decisions or chance points. Each branch (or possibility) commences at a node.

124 When people make decisions, they often use many of the processes that a Decision Tree “forces” them to use. These include (a) put things in writing, (b) group “like” things together, (c) write down possible advantages and disadvantages of decision (objectives) (d) record systematically the probability of each result or chance factor occurring –

125 The above processes provide the following advantages: Planners can (a) reduce chance of forgetting, (b) find more possibilities, (c) avoid choosingbetween possibilities until they give more time to “find possibilities”, (d) perceive interrelationships and patterns more easily, (e) discuss the situation with other people more-easily, (f) evaluate better the results and consequences because of the probabilities recorded.

126 When people make important decisions they do many things which Decision Trees help them to do. Thus, they help ensure that people use many of the good approaches to decision making.

127 However people still have difficulties in drawing Decision Trees and they avoid using them. This behaviour occurs because the discipline of forcing detailed consideration of the many complexities of a problem does not necessarily prove easy or enjoyable. It requires hard thinking work and, not unexpectedly, many people avoid this task. Besides, the delay in making a decision means the tension of “indecision” continues and people often find the temptation to get it all over and finished too great. Lack of experience in drawing Trees also leads people to avoid using them.

128 To draw a Decision Tree Planners should clearly identify the problem or decision under consideration (preferably in writing). Further, they should distinguish between (a) find possibilities and (b) evaluate them.

 

 

 

DRAWING DECISION TREES

129 The following series of steps should help to find branches for Trees: (a) Step 1: List possibilities in writing. (b) Step 2: Look for, and list (i) any common element found in two or more of the possibilities and/or (ii) any important elements that rate as part of one of the possibilities (or common elements) listed. Consider the “correctness” of the major words used for the purpose of the Tree. (c) Step 3: Look for, and record, “logical opposites” to possibilities, part-possibilities, or common elements. (d) Step 4: Re-write the possibilities in some type of Tree form and put like things together. (e) Step 5: Take each node and consider the branches recorded. Try to define all the Branches that will describe completely all possibilities from the particular node or fork of the branch. Re-draw the Tree, if necessary. (f) Step 6: Use a branch called “other” where (i) the planning has not found all possibilities from the one node and (ii) further time on this objective at that particular stage does not seem worthwhile; (g) Step 7: Consider the start of the Decision Tree. Try “move to the left”, i.e. find the “logical opposites” to the first branch.

130 A detailed example illustrates the steps. This example shows such aspects as: (a) how to use “logical opposites”, (b) some of the stages in the construction of large and complex Trees, (c) the trap of drawing inconsistent Trees, and (e) the need to define terms carefully (e.g. what people to include in the term “Buyer”)

131 Planners need to “go up the hierarchy” to check whether they have chosen the right problem.

132 The notes also discuss the following general points: (a) check that the Tree does not mix up two or more trees, (b) do not expect to draw the final Tree with the first attempt, (c) avoid extending the Tree along one horizontal branch only, (d) group similar possibilities together to avoid too many branches, (e) select the largest concept for a branch, (f) check which type of branch (decision or chance) should follow a previous branch, (g) check that two branches from the one node do not have a sequential, rather than “logical opposite”, relationship, (h) remember that the method of implementing some decisions rates as essential to consider when assessing possible results, (i) use groupings of branches which offer significant and critical classifications (e.g. break-even), (j) consider putting results branches from the one node in order of “goodness” where an objective order exists.

133 Universal Decision Trees exist and the notes show and explain one. Such Universal Trees have a general form that Planners can adapt for a variety of problems. The notes show one possibility.

 

 

 


[1] Readers familiar with the Hierarchy of Objectives should note that this step equals “going up” the Hierarchy. This approach involves asking the question: what higher-level objective (preferably use next-higher-level objective) does the objective under consideration aim to assist? Sometimes the answer obtained suggests that drawing a Tree will prove unnecessary because the Planner should abandon the objective. This type of thinking should have taken place earlier. However, even if it did, a repeat can prove useful at this stage because the problem may now have become clearer.

 

[2] This approach really belongs to Step 3 but Planners should not use the seven-step procedure too rigidly.

 

[3] These two areas really need a different classification to link with “Never Bought Product ‘X’”. The following shows one possibility - Considered: (a) No Brands, (b) One Brand Only, (c) More than One Brand.

 

Drawing Decision Trees

Introduction

1 These notes aim to help people draw Decision Trees. They describe a Decision Tree and illustrate it. The notes also list what many people do when they make decisions and point out that a Decision Tree uses most of the ideas in a written and systematic way.

2 The notes discuss a detailed step-by-step approach to drawing Decision Trees and discuss a number of general points that should help the Reader to “grow better Trees”.

 

 

Decision Making & Objective Setting

3 These notes define decision making as – a conscious thinking process aimed at choosing between two or more things. A particular class of decision making involves objective setting: a conscious thinking process aimed at choosing between two or more objectives.

4 Planning involves two stages: (a) find objectives and (b) choose between objectives, and objective setting fits into the second stage of planning.

5 Decision Trees aim not only to help make decisions (choose objectives) but also to help find possibilities. Thus, it provides a planning technique that aims to help both stages

 

 A Description of a Decision Tree

6 A Decision Tree provides a method of depicting possible decisions and their possible outcomes in a logical arrangement and sequence. The following diagram shows one way of drawing a Decision Tree –

 

 

7 Decision Trees aim to group like factors together and to display possibilities in a logical and written arrangement. Usually they start with different branches (one for each possible decision). Each of these decision branches ends in a node. From each node come branches that show possible outcomes/results (actions by others) and/or consequences.

8 The outcomes of decisions include competitive moves as well as chance factors.

9 Example. A decision to market an improved product at a higher price may lead one or more opposition companies to decrease the price for their unimproved product.

10 A Tree aims to show “every” possibility at each decision point. However, this approach could mean a very large number of branches from some points. To avoid this situation occurring, people can group like possibilities together, until the number becomes manageable.

11 Planners can label one branch as “other”. In general, Planners should avoid this approach since planners use a Decision Tree to help search for additional answers to a problem. The use of “other” makes it simple to stop searching.

12 While some articles use the sloping branches approach used in the above Tree, the “square” approach used in later diagrams allows people to fit more detail into a given area.
What people normally do when they make decisions as compared with the drawing of a decision tree

13 People make many decisions within their own mind without ever writing things down or trying to draw any type of structure such as a Decision Tree. A consideration of the various things people do will help review the desirability, or otherwise, of using Decision Trees.

14 People often do not separate clearly the two major stages of planning (find objectives and choose between them). Sometimes this approach leads people to choose between possibilities before they give adequate time to the process of finding further possibilities.

15 Experienced people often use this distinction in their own thinking. Decision Trees help to make thetask of finding possibilities a separate stage. Thus, the technique helps to keep people working on – find possibilities, i.e. people avoid the trap of moving too quickly into the stage of choosing between possibilities.

16 Many people find it useful to write down possibilities. This approach means they cannot forget them temporarily and reading over the list often brings others to mind. Such a list also helps to separate possible actions from possible results.

17 Some classification of the possibilities in the list by grouping things that have some common elements will often help to find further possibilities. It may also help to evaluate some of the possibilities.

18 Another approach aims to evaluate possibilities more systematically. One method states: select a number of the likely possibilities and write down the advantages and disadvantages of each. This approach formalises what people often do in their own mind.

19 Sometimes people gain help in finding more possibilities by seeking someone else’s opinion. This approach means that one person must communicate ideas to another person. To help do so and make better use of the Adviser’s time, some people find it better to write down the points they have in their mind in some systematic and logical way.

20 With most complex problems, people find difficulties keeping in their head the many factor involved and their interrelationships. It does not prove easy to take each likely possibility and think about likely results. Some method of writing the material down soon becomes essential if people want to explore systematically the possibilities.

21 People need help to think more clearly about the possibilities of various results occurring. Most people do so in their thinking when they classify various possibilities as “very likely” or “unlikely” etc. However, these terms rate as rather loose and people should use more specific quantitative terms. Instead of rating the probability of outcome “X” occurring as “very high” or “very low” or “about average” etc., people should use quantitative terms with reference to factors divided into 100 parts.

22 Thus if something has a very low probability, people can say it has one chance in 100 of occurring. In this case, the number 0.01 describes the probability. The probability number 0.50 refers to something that has a 50/50 chance of occurring. Something with a very high probability of occurring would use the figure of 0.95 or 0.99.

23 Example A probability of 0.99 might occur when one action involves sacking someone. In some cases, everybody in Management might believe that a l00 probability exists that a strike will occur as a consequence of the sacking.

24 All the above things can occur in a person’s mind but this approach has disadvantages. First, forgetting takes place. Second, often people only see patterns and relationships between factors when they write them down.

25 The above section aims to show that a Decision Tree does not rate as very different from what happens in real life – for important decisions.

 

Why do people (a) have difficulty in drawing, and (b) avoid using decision trees?

26 Often people find difficulties in drawing Decision Trees. However, they would find that if they considered the same problem in their own minds they would have just as much difficulty.

27 In actual practice, they might appear to have less difficulty in using their mind. However probably this situation happens because they exclude some of the factors and their interrelationships from consideration.

28 Decision Trees rate as a useful tool for Executives. However, they encourage people to spend moretime on thinking. However, thinking involves hard work and therefore people often try to avoid it. In addition, people get easily distracted from thinking about the one problem. Further, they always have the overall excuse: “I haven’t got time”.

29 To discipline oneself to draw a Decision Tree and look at all the possible outcomes will not prove easy. In any case, it delays the actual making of a decision and most people find it very satisfying and relieving (in the short term) that they have made a decision. It means that they do not have to worry about the problem any longer. Most people put aside the possibility that they have made a wrong decision and they could have many more worries in the future. The temptation always exists to risk that their decision will not prove wrong.

30 Further, it will prove difficult to depict some situations in the form of a Decision Tree. Sometimes some type of table or graph will provide a better arrangement of some types of information involved in the decision- making process.

31 In addition, people need some skill in drawing Decision Trees and disciplining their mind to think in a particular way. People will find that drawing their first ten or twenty Decision Trees will prove more difficult and a lot more time consuming than drawing Trees when relatively skilled.

32 Thus people need to gain experience and practice in drawing Decision Trees in order to make it likely they will use them.

33 Initial attempts to draw a Decision Tree often prove rather difficult because the selection of all the possibilities and their logical grouping will prove difficult. The following points which should help in this task.

 

Some General Points

34 Before starting a Tree, Planners should identify clearly the problem or decision under consideration and, preferably, put it in writing.

35 In the early stages Planners should distinguish clearly between (a) possibilities and (b) the evaluation of them (i.e. possible results) Thus, initially, Planners should just identify possible objectives without trying to evaluate them. Evaluation often encourages people not to record possibilities that sound “bad” or “silly”. Since the Tree approach aims to record all possibilities, such an early evaluation would reduce the chance of achieving this aim.

 

A suggested procedure for finding Branches for Decision Trees

36 The steps listed below apply to finding branches for both possible objectives and possible results. In future, these notes use the word “possibilities” to apply to both objectives and results. However initially they emphasise possible objectives. Most examples refer to objectives –

37 After the initial listing of the steps, the notes discuss each step and give an example to illustrate how to use each step.

38 The steps do not necessarily follow strictly one after the other. People can use any of the steps for thinking of objectives at different times during the “growing” of a Tree.

39 Step 1 – List possibilities in writing.

40 Step 2. Look for, and list: (a) any common element found in two or more of the possibilities and/or (b) any important elements which rate as part of one of the possibilities (or common elements) listed. Consider the “correctness” of the major words used for the purpose of the Tree.

41 Step 3 Look for, and record, “logical opposites” to either possibilities, part-possibilities, or common elements. (Later sections explain the meaning of logical opposites.)

42 Step 4. Re-write the possibilities in some type of Tree form, putting like things together.

43 Step 5 Take each node and consider the branches recorded that follow it. Try to define all the branches that will describe completely all possibilities from the particular node or fork of the branch. Re-draw the Tree, if necessary.

44 Step 6. Use a branch called “other” where (a) you have not found all possibilities from the one node and (b) spending further time at that stage does not seem worthwhile.

45 Step 7. Consider the start of the Decision Tree and try “moving to the left”, i.e. find the “logical opposites” to the first branch.[1]

46 Step 8. Repeat Steps 2 to 6 and record any new possibilities found (i.e. Step 1).

47 It proves difficult to know just how long to continue this process. It will become a matter of judgement by Planners. Probably it should depend on the (apparent) completeness of the Tree and the importance of the end objective for which the Planner wants to draw the Tree.

48 The following paragraphs give more details on the steps described above.

Step 1: List possibilities in writing

49 Planners can (and should) carry out this step at any time during the drawing of a Tree. People forget possibilities quite easily; thus if a possibility comes to mind, record it.

50 Planners do not necessarily have to put any possibility in the correct place on the Tree: Planners who cannot find the right place easily should record it on the side of the page or on a separate page.

51 After Planners use each possibility (not recorded on the Tree), they should show themselves that they have done so by drawing a line through it. Some such systematic procedure should ensure that they have used all possibilities on the Tree or they have discarded them.

52 Example. To make these steps understandable the following sections include an actual example. Each step will show some aspect of the example.

53 The example has the objective “Increase sales of Product X’”. Step 1 produces the following result:

(a) Give Sales Representatives more training.

(b) Look for more Distributors to take on the product.

(c) Put on more Sales Representatives.

(d) Improve the product.

(e) Drop the price.

(f) Improve the package.

(g) Alter the Product’s name.

(h) Find other uses for the product.

(i) Collect opinions from people who buy it and find out why they buy it.

54 Many more possibilities exist but the above provides sufficient to make the example useful.

55 Step 2: Look for, and list: (a) any common element found in two or more of the possibilities and/or (b) any important elements which rate as part of one of the possibilities (or common elements) listed. Consider the “correctness” of the major words used for the purpose of the Tree.

56 One approach to finding a common element involves looking for possibilities that include the same word.

57 Example. In the above example, two possibilities include the term Sales Representatives. Another common element exists - the first three possibilities and the last one all relate to people. The other possibilities all relate to something other than people.

58 Another common element exists between “Improve the product” and “Find other uses for it”. Both relate to how people use the product.

59 Product and Package have a common element in that they both refer to physical aspects of the products. Similarly, Name and Price both refer to intangible aspects.

60 The example proves just as fruitful in looking for important elements in the original possibilities listed and some of the added later ones.

61 Example. The common element “people” contains important part elements. For example, people can include the classes Buyers or Sellers. (The Sellers include both Sales Representatives and Distributors, listed in the possibilities.) To become more useful, Planners could use the classification Buyers of Product “X” and Sellers of Product “X”. However, people have other classifications. A number of people exist who do not buy Product “X”. This thought produces a better classification – Sellers and Non-Sellers and Buyers and Non-Buyers.

62 This idea leads to the consideration of whether people exist who could sell the product and at present do not. Similarly, do people exist who rate as Non-Buyers and could become Buyers? (This point rates as the central point of the whole problem.)

63 if Planners consider the possibility of collecting opinions from people who have bought “X”, logically they should look at the opposite[2] of this group, i.e. people who have not bought the product. Planners can easily move down the sub-classification to (a) the non-buyers who considered buying it and (h) those non- buyers who have never considered buying it. Those who have never considered buying “X” divide into those who have (a) heard of the product and (b) not heard of the product.

 

 

64 Further thoughts indicate that buyers of the product also have two sub-classes: Buyers of our brand of Product and Buyers of other brands of the Product.

65 The following diagram shows the above possibilities.

Possible Actions to Take Concerning Various Factors \

Diagram 1

Diagram 2

People

 

 

Diagram 3

 

66 At this stage, the above diagrams show three separate Trees. It seems a better plan to approach the problem this way, i.e. record what does fit together quite readily and then look for combinations.

67 Example. Planners can easily combine diagrams (I) and (2) by tacking (I) on to the top line of (2). The use of diagram (3) will mean a re-writing of the top part of Tree (2). A later diagram does so.

Step 3: Look for, and record, “logical opposites” to possibilities, part-possibilities, or common elements.

68 The words “logical opposite” mean more than the opposite of something, although this term includes ordinary opposites.

69 Examples. “Up” rates as the opposite to “down”, However if Planners seek “logical opposites” then going sideways would qualify, even though “side” does not rate as the ordinary opposite of “up”. The “logical opposite” (and opposite) of “move to the left” would say “move to the right”, but “move forward” or “move back” would rate as “logical opposites” as well. In addition, the logical opposite of “remain still” exists.

70 Finding “logical opposites” requires some imagination but with a little experience and practice the idea will prove easy to use. Further, many situations exist in which common “logical opposites” often occur. These opposites will help to draw a Tree for such situations.

 

 

71 Example. In the example, consider the possibility of - add to the number of Sales Representatives (or Distributors). The common or universal “logical opposites” rate as -
“increase”, “no change”, and “decrease” – with respect to the number of people concerned. The same “logical opposites” apply to price.

72 In considering “Find other uses for the Product”, the two-part classification comes to mind – Present use, Possible uses. Possible uses means future uses. This stage does not include the “logical opposite” – consider past uses which today’s consumer do not use. Probably this idea will not prove very useful but it could happen that a Company could try to revitalise a forgotten past use (perhaps after amendment) and gain acceptance (e.g. when should someone next try to re-introduce the Hula Hoop or the Yo-yo?)

 

Step 4: Re-write the possibilities in some type of tree form – putting like things together.

73 The Tree on the next page rates as appropriate. Readers should understand that this topic equals a very large one (and a large Tree) and it does not prove easy to draw.

74 The Tree on the next page merely sets out various possibilities in one systematic and logical manner. It does not specify what action to take, simply that someone could take some action with the various items identified. The specific aspects identified such as “increase our Distributors or our Sales Representatives or the Price” rate as the only specific actions listed. The rest merely identify areas.

75 The actual things that someone could do in the specific areas listed need consideration. The following possibility (already identified) exists: “collect opinions from people who buy the product and find out why they buy it”. The above analysis has shown that a number of other people exist from whom it may prove worthwhile to collect opinions. It may prove desirable to draw a Tree for this particular area. At all events, it will provide another example.

76 In the actual diagram above, the four branches: “Non-Sellers”, “Sellers of Product ‘X’”, “Buyers of Product ‘Y’”, and “Non-Buyers of Product ‘X’” do not rate as mutually exclusive. Someone could fit into the classes of both a non-seller of Product “X” and a buyer. A number of other pairs such as this pair exist. These notes ignore this problem; but will consider the actual problem of looking at buyers from whom to collect opinions. The first attempt at looking at the aspect more logically gives the following diagram.

 

 

 

 

 

 

 

 

Diagram 4

Take

Action

With

Res-

pect

to -

Diagram 5

77 However even the above does not rate as completely consistent because a buyer of Product “X” who has never heard of Product “X” cannot exist. Therefore, the Tree needs re-drawing – as occurs later.

78 At this stage, an example of a different approach should prove useful. Sometimes a diagram will help to clarify a person’s thinking better than a Tree. The following diagram aimed to help see what possibilities exist. It shows the relationship between (a) Buyers/Non- Buyers and (b) people who have considered the various brands of the products or have not considered one or more of them. The central part of the diagram shows whether the combination can, or cannot, occur.

Diagram 6

 

Considered Brand Bought

Considered Brand Bought and Others

Never Considered the Product

Buyers of Product “X” Brand “A”

Must occur

Possible

Impossible

Buyers of Product “X” Other Brands

Must occur

Possible

Impossible

Never bought Product “X”

Impossible[3]

Impossible3

Possible

DRAWING DECISION TREES

79 With Diagram 6 completed it proves easier to re-draw the Tree to get a classification which does not overlap in its categories. Diagram 7 shows this tree.

Diagram 7

 

80 Even the above diagram does not allow for “Buyer of both Product “X” and other brands”.

81 A Planner could add more detail to the above diagram in the area marked (A) by using the following:

Diagram 8

 

 

 

82 A Planner could sub divide the branch “Never bought Product “X” in a different way as shown in the following diagram

 Diagram 9

83 These notes have not defined the word “Buyer” as used in the diagram above. A Planner would introduce a further complexity by talking about various levels of buying strength.

84 Example. Regular buyers over the last two years, buyers who have just started buying the product, buyers who have just changed from one brand to our brand, and so on. The possible complexities will become almost infinite. The desirability of taking them to this extreme would depend on the activities involved and the importance of the whole objective in this area in particular.

85 At this stage Readers may well ask – Just where will this approach lead? and Will it prove useful?

86 First, it does identify the people from whom investigators can seek opinions. However – could a Planner have reached the classifications in the latter diagrams and understood some of their relationships by using a better/shorter method? Would a less-complete classification prove just as useful?

87 The next stage would become – decide what opinions to collect. The following question shows the scope: “Why did you buy, or did not buy, the particular product and/or particular brand of the product?”

88 Once someone has identified the classifications then someone can aim to evaluate the possible results of the various types of questions to the various classes of people. Perhaps researchers will not need opinions in a particular market research from people who have never considered buying the product. Early in the interview once an Interviewer identifies a Respondent as belonging to this particular category, they would ask no more questions. It can also occur that certain types of questions will prove useless to ask of people who have never considered “our Brand”.

89 With some grouping of these classes of people and the addition of why they did (or did not) buy, it will prove easier to look at the possibilities of various marketing activities having different effects.

90 Example Investgators could look at the activity of altering the price from the viewpoint of these various classes of people. They could try to assess the likelihood of them buying more, less, or having little effect - in relation to a price increase, or decrease or no change.

91 The possibility “Collect opinions, etc.” has proved an extremely complex one. However, this type of situation often happens in an organisation: someone asks someone to do something that rates as very broad and unless someone works out the details often what people do proves wasted. Quite often, the actual amount of work exceeds what the original person suggesting the approach envisaged. Sometimes people carry out the broad objective quite poorly. People should not feel surprised at the useless results that often occur.

 

Step 7: Consider the start of the Decision Tree and try “moving to the left”, i.e. find the “logical opposites” to the first branch.

92 The discussion of the example has illustrated the processes of Step 5 and Step 6 but it would take a great deal more effort to draw the whole Tree for the objective of the example and it would cover a very large sheet of paper.

93 Example. It will prove simple to find some “logical opposites” of the objective - “Increase sales of Product ‘X’”, The possibilities include: (a) maintain the present sales, (b) decrease the sales, or (c) increase them. The area of decreasing sales has an end to it in the sense that the objective could state – achieve no sales. This objective equals deleting the product from the company’s operations. Although Planners may have considered this objective, they should keep it well to the fore since it will prove easy to forget or overlook this “obvious” possibility.

94 Investigators could first aim to establish the likelihood that sales of the product can increase. They can compare two objectives – (a) spend money for market research on finding out why people buy and do not buy the product versus (b) ask questions about the real need which the product has in the past satisfied and will anything change (reduce) this need and will other products satisfy it better,

General Points

95 Apart from the steps discussed above the following general points should help in drawing Decision Trees.

96 Check that you aim to draw only one Tree at a time. Sometimes a number of problems will exist at the same time and interact. A Tree may prove difficult to draw because no one has established the separating point between the problems. Planners should do so.

97 Do not try too hard to fit all the possibilities into the initial Tree. Planners who have difficulty finding a place for a possibility in a Tree should write it down on one side of the page. At a later stage, it may prove much easier to fit it into the structure – because the drawing of the Tree has established a better understanding of the whole situation. Eventually it must prove possible to fit it in or feel sure that it belongs to a different problem.

98 With complex problems, it will often help to draw “small” Trees for different parts of the problem and then try to put them together.

99 Develop the Tree from all nodes on each vertical level – rather than extending the Tree on one horizontal level much further than at other levels.

100 In drawing the possibilities from each node, group together similar possibilities. Probably no node should have more than about five or six branches coming from it.

101 Use the largest concept or idea in deciding on a branch.

102 Examples. Profitability rates as a larger concept that sales since the interaction between sales and expenses gives profit. “Go home” (for a person already at work) gives a “smaller” idea than “Leave work”. In this case “Go home” rates as the smaller concept since “Go home” means that the person must “Leave work”, but the person could also “Go to the races” or “Go out to lunch”. The larger concept “Leave work” covers all the possibilities but “Leave work” does not necessarily mean that the person must “Go home”.

103 Decide when another decision node(as opposed to a chance node) should follow an originaldecision node.Similarly decide when a chance node should follow another chance node.

104 Some two-part decisions only help to group possibilities logically. Thus some things to do must follow a “do something” (versus “do nothing”) node.

105 A variation of this point states – some possible decisions have a sequential link; they can never exist as alternatives.

106 Example. Probably the objectives - “Phone the Supervisor” and “Ignore the different instructions given” rate as sequential. The following diagram shows an appropriate Tree of the possibilities.

Diagram 10.

107 A slightly-different form might improve the tree:

Diagram 11

108 The diagram in a different form shows more clearly that the action of informing someone rates as independent of the decision to accept or reject different instructions.

In the context of - I will inform my Supervisor, I will not carry out (i.e. I reject) his instructions, the term ignore does not fit. Usually people ignore instructions but do not tell their Supervisor that they intend to ignore the instructions. The term “passive resistance” describes this situation

109 Another reason exists which supports the idea that a chance factor should not necessarily follow a decision node. Sometimes the particular methods used to implement a decision will prove almost as significant as the decision. In this case, the reactions to the decision should really take into account the means of implementing the decision.

110 This point holds true particularly in personnel problems. However with broader scale objectives (e.g. building a new factory or launch a new product) the implementation rates as important, but often it will prove useful to assume that the implementation will proceed with average efficiency. Where implementation largely depends on one or a few factors (e.g. the persuasiveness of one person) then Planners should link the methods with the implementation of the decision.

111 Look for critical and significant possibilitieswhen grouping and arranging possibilities.

112 Example. In looking at consequences of a decision, a critical point often becomes whether the result will reach a break-even point (where income equals expenditure). Thus, the three branches from a consequence or chance node might show (a) makes a profit, (b) breaks even, (c) makes a loss.

113 One interesting example occurs with the result – “effect on productivity”. It can involve increase, no change, and decrease. However, sometimes Planners should look carefully at a special point of decrease, i.e. nil productivity. In certain industrial situations, this point means that a strike has resulted.

114 Recording branches in order on some scale.

115 Examples. Record good, bad, or neutral results in order: Good, neutral, bad (or bad, neutral, good). High Profit, Medium Profit, Break Even, Loss shows another set of possibilities arranged in order

 116 Sometimes Planners should put all the good possibilities at a higher (branch) level than poor ones. This point assumes that Planners can evaluate easily the “goodness” of a branch. In some cases, it can (e.g. profit versus loss); in others it cannot. (This suggestion also tends to oppose the previous suggestion of not evaluating possibilities but sometimes the evaluation rates as obvious and not a matter of personal opinion.)

117 Remember that Planners draw a Tree to help make decisions. The provision of many branches or their elaboration can waste time and money. However once someone makes particular decisions then it will prove appropriate to elaborate (draw another Tree) on the particular branches chosen.

 

Universal Trees

118 Some Trees exist which rate as appropriate for a number of problems. Their structure may need adaptation for particular problems; they will certainly need translation.

 

 

 

 

 

 

 

 

119 The following shows one example.

“Universal” Decision Tree Form for Decisions

120 This Tree could not show easily all the possible consequences that might arise from any one decision since a large number of decisions exist. However, Planners could select the four or six “best” and draw the consequences of these decisions.

121 For Planners who select the decision “collect further information”, methods of continuing or abandoning the objective will not prove appropriate until someone obtains the “further information”. One method of continuing the Tree would involve listing the likely broad results that might come from the further investigation. Then it could show decisions on how to continue or abandon. However probably this approach would prove inappropriate unless the whole matter rated as so urgent that Planners should spend time in planning to enable a quick start on continuing the problem immediately they obtained “further information”. Unless this situation existed, it would prove better to defer further thinking until someone obtained the “further information”.

 

Summary

122 A Decision Tree shows possible decisions and their possible outcomes written out in a logical arrangement and sequence. It aims to help both stages of planning, i.e. (a) find possibilities (objectives) and (b) choose between possibilities. These notes call the second stage of planning “decision making”.

123 In Decision Trees, Branches show possible actions and consequences. Nodes (or forks) show decisions or chance points. Each branch (or possibility) commences at a node.

124 When people make decisions, they often use many of the processes that a Decision Tree “forces” them to use. These include (a) put things in writing, (b) group “like” things together, (c) write down possible advantages and disadvantages of decision (objectives) (d) record systematically the probability of each result or chance factor occurring –

125 The above processes provide the following advantages: Planners can (a) reduce chance of forgetting, (b) find more possibilities, (c) avoid choosingbetween possibilities until they give more time to “find possibilities”, (d) perceive interrelationships and patterns more easily, (e) discuss the situation with other people more-easily, (f) evaluate better the results and consequences because of the probabilities recorded.

126 When people make important decisions they do many things which Decision Trees help them to do. Thus, they help ensure that people use many of the good approaches to decision making.

127 However people still have difficulties in drawing Decision Trees and they avoid using them. This behaviour occurs because the discipline of forcing detailed consideration of the many complexities of a problem does not necessarily prove easy or enjoyable. It requires hard thinking work and, not unexpectedly, many people avoid this task. Besides, the delay in making a decision means the tension of “indecision” continues and people often find the temptation to get it all over and finished too great. Lack of experience in drawing Trees also leads people to avoid using them.

128 To draw a Decision Tree Planners should clearly identify the problem or decision under consideration (preferably in writing). Further, they should distinguish between (a) find possibilities and (b) evaluate them.

 

 

 

DRAWING DECISION TREES

129 The following series of steps should help to find branches for Trees: (a) Step 1: List possibilities in writing. (b) Step 2: Look for, and list (i) any common element found in two or more of the possibilities and/or (ii) any important elements that rate as part of one of the possibilities (or common elements) listed. Consider the “correctness” of the major words used for the purpose of the Tree. (c) Step 3: Look for, and record, “logical opposites” to possibilities, part-possibilities, or common elements. (d) Step 4: Re-write the possibilities in some type of Tree form and put like things together. (e) Step 5: Take each node and consider the branches recorded. Try to define all the Branches that will describe completely all possibilities from the particular node or fork of the branch. Re-draw the Tree, if necessary. (f) Step 6: Use a branch called “other” where (i) the planning has not found all possibilities from the one node and (ii) further time on this objective at that particular stage does not seem worthwhile; (g) Step 7: Consider the start of the Decision Tree. Try “move to the left”, i.e. find the “logical opposites” to the first branch.

130 A detailed example illustrates the steps. This example shows such aspects as: (a) how to use “logical opposites”, (b) some of the stages in the construction of large and complex Trees, (c) the trap of drawing inconsistent Trees, and (e) the need to define terms carefully (e.g. what people to include in the term “Buyer”)

131 Planners need to “go up the hierarchy” to check whether they have chosen the right problem.

132 The notes also discuss the following general points: (a) check that the Tree does not mix up two or more trees, (b) do not expect to draw the final Tree with the first attempt, (c) avoid extending the Tree along one horizontal branch only, (d) group similar possibilities together to avoid too many branches, (e) select the largest concept for a branch, (f) check which type of branch (decision or chance) should follow a previous branch, (g) check that two branches from the one node do not have a sequential, rather than “logical opposite”, relationship, (h) remember that the method of implementing some decisions rates as essential to consider when assessing possible results, (i) use groupings of branches which offer significant and critical classifications (e.g. break-even), (j) consider putting results branches from the one node in order of “goodness” where an objective order exists.

133 Universal Decision Trees exist and the notes show and explain one. Such Universal Trees have a general form that Planners can adapt for a variety of problems. The notes show one possibility.

 

 

 


[1] Readers familiar with the Hierarchy of Objectives should note that this step equals “going up” the Hierarchy. This approach involves asking the question: what higher-level objective (preferably use next-higher-level objective) does the objective under consideration aim to assist? Sometimes the answer obtained suggests that drawing a Tree will prove unnecessary because the Planner should abandon the objective. This type of thinking should have taken place earlier. However, even if it did, a repeat can prove useful at this stage because the problem may now have become clearer.

 

[2] This approach really belongs to Step 3 but Planners should not use the seven-step procedure too rigidly.

 

[3] These two areas really need a different classification to link with “Never Bought Product ‘X’”. The following shows one possibility - Considered: (a) No Brands, (b) One Brand Only, (c) More than One Brand.

 

Reminders

Reminders

During a course, Attendees may learn a number of new ideas. However how many of the Attendees remember  them – let alone put them into practice. These Reminders aim to provide Organisations with some one-page pieces that they can send out to Course Members after they finish a course to remind them of the ideas. This approach also hopes that a Reminder will also encourage the Receiver to put the ideas in the reminder into practice. (25/3/10).

This file lists all the Reminders – in the order shown on these pages.

.A Reminder about the use of Competed Staff Work.

Managers can waste the time of their Subordinates by insisting that they receive material by a certain time and they do not use it for several days.

A reminder about the possibilities of saving the time of people who you ring and the people in the office of the call receiver. (27/3/10)

In any formal meeting it helps to set aside the rules and have a round-table discussion. This reminder discusses this possibility. (27/3/10)

This reminder helps Readers to remember to start all objectives with a verb and use objectives to draw hierarchies.

A reminder on some ideas that will help Senders to obtain answers to the questions they ask.

This reminder asks Readers to consider how much (if any) they  have improved the performance of each of their Subordinates. Can Managers ask their Subordinates to tell them what they have learnt from their Manager in recent months?


Readers should remember to check whether their conversation about forecasts refers to what they expect will happen or what they desire to happen. (27/3/10)

If you allow a disturbance in a group of eight people and it takes three minutes, you have wasted 24 person minutes. This reminder suggests that you take steps to stop allowing others to disturb your conferences.

ROLE PLAYING CAN ASSIST IN MORE THINGS THAN JUST SALES TRAINING.
 Should Managers insist that their Subordintates use the method that the Manager prefers to achieve a particular obkjective? Managers should back off on some occasions and let their Subordinates use the method that the Subordinate prefers. This reminder discusses this topic.  (28/3/10)
If you allow a disturbance in a group of eight people and it takes three minutes, you have wasted 24 person minutes. This reminder suggests that you take steps to stop allowing others to disturb your conferences.

ROLE PLAYING CAN ASSIST IN MORE THINGS THAN JUST SALES TRAINING.
Should Managers insist that their Subordinates use the method that the Manager prefers to achieve a particular obkjective? Managers should back off on some occasions and let their Subordinates use the method that the Subordinate prefers. This reminder discusses this topic.SUGGESTING TO YOUR BOSS THAT YOU MAKE THE DECISION

 

 


Home

This blog aims to include many notes on the subject of Management.

For my understanding of Management see the page called Management and the file called – Introduction to Management.

It will take many weeks before I can upload most of my material (13/1/12)