Software Forensics Centre Decisive and Incisive Making Decisions

Software Forensics Centre Decisive and Incisive: Making Decisions in IT and project management Prof. Darren Dalcher Software Forensics Centre School of Computing Science Middlesex University d. dalcher@mdx. ac. uk © Darren Dalcher 1

Software Forensics Centre Outline: decision making, bias, professional Ø The link between decision making and bias Ø Challenges in decision making Ø Are you susceptible to biases? Ø Framing questions Ø Dealing with uncertainty Ø Too many choices? Ø Implications Ø The way forward Ø Professional decision making? © Darren Dalcher 2

Software Forensics Centre What is bias? Ø According to the Oxford Dictionary: A bias is a predisposition or prejudice. Ø According to my son’s English book it is a strong opinion for or against an idea. Ø But are we susceptible to biases? © Darren Dalcher 3

Software Forensics Centre Would you buy this house? This charming and unusual country cottage enjoys a quiet position at the end of an idyllic country lane. Its period features and unashamed old world style make it ideal for restoration by a loving owner. The property benefits from many original characteristics which could be further enhanced to suit your personal taste. © Darren Dalcher The ‘house’ stands in a very isolated position up a rough and unmade farm track. It is sadly neglected – the roof tiles are broken, the walls and ceiling crumbling and there is an urgent need for essential repairs. Inside most of the rooms there is evidence of rising damp. The whole property requires an urgent overhaul. 4

Software Forensics Centre The starting point Most of us do not make great decisions, and a few of us are aware of this fact. - Hoch & Kunreuther Today we are going to explore this area. © Darren Dalcher 5

Software Forensics Centre Challenges in decision making Ø Information overload Ø Galloping rate of change Ø Rising uncertainty Ø Few historical precedents Ø More frequent decisions Ø More important decisions Ø Conflicting goals Ø More opportunities for miscommunication Ø Fewer opportunities to correct mistakes Ø Higher stakes (after Russo & Schemaker) © Darren Dalcher 6

Software Forensics Centre Making decisions Ø New play by Andrew Lloyd Webber, Rats Ø You just bought a ticket for £ 50, seats not marked. You lost your ticket. Would you buy another? Ø Situation 2: Going to see a play. Open wallet – lost £ 50 note. Would you still pay? Ø Typically 62% buy ticket again (having made effort) Ø 85% pay cash © Darren Dalcher 7

Software Forensics Centre Gamble 1% chance of winning £ 10, 000 50% chance of winning £ 400 Which option would you select? Expected Value 0. 01 * £ 10, 000 = £ 100 0. 50 * £ 400 = £ 200 © Darren Dalcher 8

Software Forensics Centre Bias Ø Many decisions deviate from the ‘optimal, rational’ result. The attempts to apply a rational strategy or approach are often flawed. The flaws are repeatable and can be attributed to consistent mistakes. Ø These cognitive biases reduce the effectiveness of the decision maker affecting the acquisition, analysis and interpretation phases. This in turn affects the quality (and relevance) of the decision itself. © Darren Dalcher 9

Software Forensics Centre What treatment would you prefer? Of 100 people having surgery, 10 will die during surgery, 32 will have died within one year and 66 will die within five years. Of 100 people having radiation therapy, none will die during treatment, 23 will die within one year, and 78 will die within five years. Of 100 people having surgery, 90 will survive the surgery, 68 will survive past one year, and 34 will survive through five years. Of 100 people having radiation therapy, all will survive the treatment, 77 will survive one year, and 22 will survive past five years. © Darren Dalcher 10

Software Forensics Centre Can we trust professionals? New England Journal of Medicine Ø Group of 369 boys were examined for possible tonsillectomy Ø Step 1: Panel of doctors examined all boys Decision: 45% required tonsillectomy Ø Step 2: New panel of doctors examined remaining boys (214) Decision: 46% needed tonsillectomy Ø Step 3: New panel of doctors examined 116 doubly healthy boys Decision: 44% needed tonsillectomy © Darren Dalcher 11

Software Forensics Centre Availability Ø The decision maker uses only easily available information and ignores not easily available sources of significant information. Ø Are there more seven letter words in the English language: - - - n – - - i n g Ø Consider a group of ten people: How many different committees of two? How many different committees of eight? © Darren Dalcher 12

Software Forensics Centre Availability again Ø Are there more words in the English language a. that begin with the letter ‘R’ ? b. For which ‘R’ is the third letter ? Ø Most of us can recall words that begin with ‘R’. However, thinking of words that have R as the third letter is more difficult. © Darren Dalcher 13

Software Forensics Centre Anchoring & adjustment Ø Selecting an initial value as a starting point (could be average) and then adjusting the value improperly in order to accommodate the rest of the data. Ø Provide a rapid guesstimate for: 1 * 2 * 3 * 4 * 5 * 6 * 7 * 8 = 8 * 7 * 6 * 5 * 4 * 3 * 2 * 1 = Medians typically are 512 and 2, 250 Answer: 40, 320 © Darren Dalcher 14

Software Forensics Centre Hiring a new engineer Ø A newly hired engineer for a computer firm in the City of London area has four years experience and excellent qualifications. When asked to estimate the starting salary for this employee, my secretary (knowing very little about the profession or industry) guessed an annual salary of £ 23, 000. Ø What is your estimate of the salary? © Darren Dalcher 15

Software Forensics Centre Hiring a new engineer II Ø A newly hired engineer for a computer firm in the City of London area has four years experience and excellent qualifications. When asked to estimate the starting salary for this employee, my secretary (knowing very little about the profession or industry) guessed an annual salary of £ 85, 000. Ø What is your estimate of the salary? © Darren Dalcher 16

Software Forensics Centre Some more unrelated anchoring Ø What are the last three digits of your home phone number? Ø Add 400 to the number above. Ø In what year would you guess Attila the Hun was actually defeated? Range (number + 400) Mean 400 to 599 600 to 799 800 to 999 1000 to 1199 1200 to 1399 © Darren Dalcher 629 680 789 885 988 17

Software Forensics Centre Data Presentation Ø The impact of summarised data and facts Ø Would you undertake an action that will produce a 30% increase in the annual mortality risk? Ø Would you undertake an action increasing your annual chances of death from 1 in 10, 000 to 1. 3 in 10, 000? © Darren Dalcher 18

Software Forensics Centre Buying a new item – adding features Ø You are about to buy a new car for £ 22, 134. 56 and have just been tempted to add a new radio which will bring the total price to £ 22, 428. 41. (feeling that the difference is trifling) Ø Is this a rational decision? Ø Upon reflection you can see that you are about to be charged £ 293. 85 for a radio, which may be a touch excessive… © Darren Dalcher 19

Software Forensics Centre Conservatism Ø Failure to revise estimates (by as much as they should be) in light of new and significant information. © Darren Dalcher 20

Software Forensics Centre Data saturation Ø People often reach premature conclusions on the basis of too small a sample of information while ignoring the rest of the data that is received later. © Darren Dalcher 21

Software Forensics Centre Self-fulfilling prophecies Ø Selective perception occurs when a certain outcome, interpretation or conclusion is favoured and information that does not support this view is ignored. Ø People often remember and attach higher validity to information which confirms their previously held beliefs and expectations. On the other hand they selectively ignore disconfirming information. © Darren Dalcher 22

Software Forensics Centre Ease of recall Ø Data which can be easily recalled or assessed is often used in estimates. An event is believed to occur frequently, that is with high probability, if it is easy to recall similar events. Ø Direct experience and coverage by media can play a part. © Darren Dalcher 23

Software Forensics Centre Fact-value confusion Ø Strongly held values may often be regarded and presented as objective facts which are allowed to determine choices and affect decisions. © Darren Dalcher 24

Software Forensics Centre Success/failure attribution Ø Decision makers often associate success with personal ability or exceptional skills and failure with bad luck and unfavourable circumstances. © Darren Dalcher 25

Software Forensics Centre Gamblers fallacy Ø Assumptions regarding runs of probability and sequences of events that seem to suggest that a run of poor results will be followed by an extremely successful outcome. © Darren Dalcher 26

Software Forensics Centre Habit Ø Familiarity with a particular rule, method or approach may often lead to re-utilisation of the same procedure or the selection of a similar alternative when faced with a seemingly identical problem. © Darren Dalcher 27

Software Forensics Centre Order effects Ø The order in which items are presented affects information retention in memory. Ø Typically the first piece of information presented (primacy effect) and the last one (recency effect) assume undue importance in the mind of the decision maker. © Darren Dalcher 28

Software Forensics Centre Redundancy Ø The more redundancy in the data, the more confidence people tend to place in their predictions. © Darren Dalcher 29

Software Forensics Centre Hindsight Ø People are often unable to think objectively if they receive information that an outcome has occurred and they are told to ignore this information. With hindsight outcomes that have occurred seem inevitable. Relationships and links are easier to form and predictions are often altered to reflect what we know occurred. Ø Football results Ø Lost in traffic? Ø New employee in interview © Darren Dalcher 30

Software Forensics Centre Using hindsight to sell Ø Storer Cable Communication sent the following notice to subscribers in Louisville, Kentucky. Ø “It’s not often you get good news instead of a bill, but we have got some for you. If you’ve heard all the rumours about your basic cable rate going up $10 or more a month, you can relax: It’s not going to happen! The great news is … the rate for basic cable is increasing only $2 a month” © Darren Dalcher 31

Software Forensics Centre How consistent are you? Ø You are in a shop about to buy a new watch which will cost £ 70. As you wait to pay, a friend walks past and tells you that an identical watch is available in another shop 10 minutes away for £ 40. You know that the service and reliability of the other shop are just as good as this one. Will you travel 10 minutes to save £ 30? © Darren Dalcher 32

Software Forensics Centre How consistent are you? Ø You are in a shop about to buy a new video camera which will cost £ 800. As you wait to pay, a friend walks past and tells you that an identical video camera is available in another shop 10 minutes away for £ 770. You know that the service and reliability of the other shop are just as good as this one. Will you travel 10 minutes to save £ 30? © Darren Dalcher 33

Software Forensics Centre Framing questions - obtaining information Ø Do you get headaches frequently, and if so how often? 2. 2 / week Ø Do you get headaches occasionally, and if so how often? 0. 7 / week Ø How many headache products have you tried? 1? 5? 10? 5. 2 products Ø How many headache products have you tried? 1? 2? 3? 3. 3 products • After Loftus (1975) © Darren Dalcher 34

Software Forensics Centre Framing questions – changing facts? Ø How long was the film? Ø How short was the film? 130 minutes 100 minutes Ø How tall was the player? Ø How short was the player? 79 inches 69 inches • After Harris (1973) © Darren Dalcher 35

Software Forensics Centre Framing questions: the right time? Ø Guardian survey Ø Should Britain ever join the Euro? Join: 29% Not join: 52% undecided 19% Ø Should Britain join as soon as conditions are right? Join: 46% © Darren Dalcher 36

Software Forensics Centre Dealing with uncertainty Ø Imagine you have just taken a very tough final examination. It is the end of the second semester, you feel tired and run-down, and you are not sure that you passed the exam. Failing the exam means a re-take after the summer break. Ø You have an opportunity to buy a very attractive two week holiday package at an exceptionally low price. The special offer expires tomorrow, while the exam grades will not be available until the following day. © Darren Dalcher 37

Software Forensics Centre Dealing with uncertainty Ø Would you: (a) buy the package (b) not buy the package (c) pay a £ 10 non-refundable fee to retain the right to buy the package at the special promotional price (a) 32% (b) 7% (c) 61% © Darren Dalcher 38

Software Forensics Centre Pass/fail formulations Ø Two additional formulations of the problem presented to groups of students: (a) buy (b) not buy (c) pay £ 10 © Darren Dalcher Pass 54% 16% 30% Fail 57% 12% 31% Original 32% 7% 61% 39

Software Forensics Centre Insurance/ Gamble and uncertainty Insurance Formulation Situation A: You stand a one out of a thousand chance of losing £ 1000. Situation B: You can buy insurance for £ 10 to protect you from this loss. Gamble Formulation Situation A: You stand a one out of a thousand chance of losing £ 1000. Situation B: You will lose £ 10 with certainty. © Darren Dalcher 40

Software Forensics Centre Winning or losing? Imagine I have just given you £ 200. I now offer you more, in the form of one of two options: Option 1. I will give you an additional £ 100 Option 2. I will toss a fair coin. If it lands heads, I will give you an additional £ 200; if it lands tails, I will give you no additional money. Option 1. You must give me back £ 100 Option 2. I will toss a fair coin. If it lands heads, you must give me back £ 200; if it lands tails, you may keep all the money I gave you. © Darren Dalcher 41

Software Forensics Centre Assigning values Ø Imagine the UK is preparing for an outbreak of an unusual disease that is expected to kill 600 people. Two programmes to combat the disease have been proposed. A. If prog. 1 is adopted 200 people will be saved. B. If prog. 2 is adopted, there is a one-third probability that 600 people will be saved, and a two-thirds probability that no people will be saved. A’. If prog, 1 is adopted 400 people will die. B’. If prog. 2 is adopted, there is a one-third probability that nobody will die, and a two-thirds probability that 600 people will die. © Darren Dalcher 42

Software Forensics Centre The ‘blessing’ of choice Ø Added choice tends to reduce the value of any single item Ø When offered $1. 50 for participating in a survey or a pen (they were told is worth $2), 75% of subjects chose the pen Ø When offered $1. 50 or a choice between the same pen and a set of cheaper felt-tipped pens, the majority went for the money (less than 50% chose any pen) © Darren Dalcher 43

Software Forensics Centre The winner’s curse Ø You are in a foreign country and meet a merchant who is selling you a very attractive diamond/rug… Ø You have purchased a few of these … in your life, but are far from being an expert. Ø After some discussion you make what you are sure is a very low offer. The merchant immediately accepts. How do you feel? Ø Why would you voluntarily make an offer that you do not want accepted? © Darren Dalcher 44

Software Forensics Centre Regret Ø The problem is that you have less information than the other side. Ø Buying a house Ø Your agent tells you of another party about to make an offer. You were going to offer £ 200 K, but now offer £ 220 K which is accepted on the spot. How do you feel? © Darren Dalcher 45

Software Forensics Centre A partial model of decision making © Darren Dalcher 46

Software Forensics Centre Levels of bounded rationality © Darren Dalcher 47

Software Forensics Centre About biases “Get your facts first, and then you can distort them as much as you please” ― Mark Twain © Darren Dalcher 48

Software Forensics Centre Implications for our profession Ø Sw. E Bo. K not enough Ø Need to know about decision making? Ø Decision Making a key competence? Ø Licensing Professionals Ø Responsibility and liability of engineers Ø Do we need to teach decision making? Ø Ethical issue – Objective DM? Ø CPD in DM? Ø Researchers: implications on how we ask questions? © Darren Dalcher 49

Software Forensics Centre The way forward Ø Think what is outside your view (assumptions) Ø Be aware of anchoring Ø View decisions from multiple perspectives Ø Work with multiple sources of information Ø Look at the complementary side Ø Provide a range, not an isolated value Ø Challenge your views and values Ø Remain open to new (& conflicting) information Ø Know the limits of your information (beyond facts) – confidence levels Ø Remember we are all human! © Darren Dalcher 50

Software Forensics Centre Summary Ø The good Ø The bad Ø And the professional © Darren Dalcher 51

Software Forensics Centre Dealing with people Ø The biggest winner ever in the Spanish lottery was asked how he chose his winning ticket. Ø He said he looked around for a ticket ending in the number 48 as he knew it would win Ø Asked how he knew, he explained that… Ø For seven nights he had a dream about the number seven, and as 7 * 7 is 48, he just went looking for the winner… © Darren Dalcher 52