Cognitive Biases in Engineering Organizations - Craft Conf

Craft Conf – April 29th, 2016 @jonathanklein Jonathan Klein in
Engineering Organizations Cognitive Biases

Talk Resources (Slides/Links) jkle.in/craft

jkle.in/craft @jonathanklein 3

jkle.in/craft @jonathanklein Who Am I? 3

jkle.in/craft @jonathanklein Who Am I? • Director of Engineering at
Attend 3

Attend • Co-Author of Hound (a code search tool) 3

Attend • Co-Author of Hound (a code search tool) • I like ﬁguring out why we believe what we believe 3

@jonathanklein jkle.in/craft Cognitive Bias: A deviation in judgement where inferences
may be illogical 5

@jonathanklein jkle.in/craft System 1 Operates automatically and quickly, with little
or no effort and no sense of voluntary control 6

@jonathanklein jkle.in/craft System 2 Allocates attention to the effortful mental
activities that demand it, including complex computations 7

@jonathanklein jkle.in/craft System 1 Operates automatically and quickly, with little
or no effort and no sense of voluntary control 8

@jonathanklein jkle.in/craft Daniel Kahneman Focuses on Problematic Biases 11 Gary
Klein Focuses on Helpful Biases (NDM)

“Although DK is thrilled by the remarkable intuitive skills of
experts that GK and others have described, he also takes considerable pleasure in demonstrations of human folly and in the comeuppance of overconﬁdent pseudo-experts. For his part, GK recognizes that formal procedures and algorithms sometimes outdo human judgment, but he enjoys hearing about cases in which the bureaucratization of decision making fails.” Conditions for Intuitive Expertise A Failure to Disagree

@jonathanklein jkle.in/craft 5-10% of videos were upside-down 14

“It was designed for right-handed users, but phones are usually
rotated 180 degrees when held in left hands.” 15

jkle.in/craft @jonathanklein • Projection Bias • Planning Fallacy • Bandwagon
Effect • Sunk Cost Fallacy • Hyperbolic Discounting • Fundamental Attribution Error 16

Part 1: A Fictional Story

@jonathanklein jkle.in/craft Sales Guy 19

@jonathanklein jkle.in/craft Engineering Manager 20

@jonathanklein jkle.in/craft Software Engineer 21

@jonathanklein jkle.in/craft CEO 22

@jonathanklein jkle.in/craft Sales Guy Sells a feature that doesn’t exist
yet 24

@jonathanklein jkle.in/craft Engineering Manager ROI is positive at < 6
weeks, thinks we can do it in 4 25

@jonathanklein jkle.in/craft Software Engineer “Yeah, we can do that” 26

@jonathanklein jkle.in/craft Two weeks of development… 27

@jonathanklein jkle.in/craft Software Engineer “Oh sh*t, we need 8 more
weeks” 28

@jonathanklein jkle.in/craft Engineering Manager “Let’s push forward, and hope we
beat expectations” 29

@jonathanklein jkle.in/craft Software Engineer “Okay, we’ll cut some corners” 30

@jonathanklein jkle.in/craft Six more weeks go by… 31

@jonathanklein jkle.in/craft CEO “Who do we need to ﬁre?” 33

This could have been avoided 34

Deconstructing the Story

@jonathanklein jkle.in/craft Sales Guy Sells a feature that doesn’t exist
yet 36

@jonathanklein jkle.in/craft Projection Bias One thinks that others have the
same priority, attitude or belief that one harbors oneself, even if this is unlikely to be the case. 37

same priority, attitude or belief that one harbors oneself, even if this is unlikely to be the case. Solution 37

same priority, attitude or belief that one harbors oneself, even if this is unlikely to be the case. Solution Empathy 37

same priority, attitude or belief that one harbors oneself, even if this is unlikely to be the case. Solution Empathy Have people switch roles 37

@jonathanklein jkle.in/craft Planning Fallacy The tendency for people to underestimate
how long they will need to complete a task, even when they have experience of similar tasks over-running 39

how long they will need to complete a task, even when they have experience of similar tasks over-running Solution 39

how long they will need to complete a task, even when they have experience of similar tasks over-running Solution Prototyping 39

how long they will need to complete a task, even when they have experience of similar tasks over-running Solution Prototyping Use information from similar ventures 39

how long they will need to complete a task, even when they have experience of similar tasks over-running Solution Prototyping Use information from similar ventures “Planning poker” exercise 39

@jonathanklein jkle.in/craft Planning Poker Cards 40

@jonathanklein jkle.in/craft 41

@jonathanklein jkle.in/craft This works because it changes the reference point
42

@jonathanklein jkle.in/craft “Reference Class Forecasting” 43

@jonathanklein jkle.in/craft Story Point Best Practices 44

@jonathanklein jkle.in/craft Story Point Best Practices Points != Time 44

@jonathanklein jkle.in/craft Story Point Best Practices Points != Time Points
Don’t Translate Across Teams 44

@jonathanklein jkle.in/craft Story Point Best Practices Points != Time Points
Don’t Translate Across Teams Points Don’t Work For Huge Projects 44

@jonathanklein jkle.in/craft Software Engineer “Yeah, we can do that” 47

@jonathanklein jkle.in/craft Bandwagon Effect The rate of uptake of beliefs,
ideas, fads and trends increases the more that they have already been adopted by others 48

ideas, fads and trends increases the more that they have already been adopted by others Solution 48

ideas, fads and trends increases the more that they have already been adopted by others Solution Make decisions asynchronously 48

ideas, fads and trends increases the more that they have already been adopted by others Solution Make decisions asynchronously Ask for conﬁdential feedback 48

ideas, fads and trends increases the more that they have already been adopted by others Solution Make decisions asynchronously Ask for conﬁdential feedback Don’t prime the estimate 48

ideas, fads and trends increases the more that they have already been adopted by others Solution Make decisions asynchronously Ask for conﬁdential feedback Don’t prime the estimate “Planning poker” exercise 48

@jonathanklein jkle.in/craft Software Engineer “Oh sh*t, we need 8 more
weeks” 49

@jonathanklein jkle.in/craft Engineering Manager “Let’s push forward, and hope we
beat expectations” 50

@jonathanklein jkle.in/craft Sunk Cost Fallacy A sunk cost is a
retrospective (past) cost that has already been incurred and cannot be recovered 51

retrospective (past) cost that has already been incurred and cannot be recovered Solution 51

retrospective (past) cost that has already been incurred and cannot be recovered Solution Do some math 51

retrospective (past) cost that has already been incurred and cannot be recovered Solution Do some math At every point in the project, the ROI should still be positive 51

@jonathanklein jkle.in/craft 52 Return on Investment

@jonathanklein jkle.in/craft 53 Return on Investment

@jonathanklein jkle.in/craft 53 Return on Investment Project Completed

@jonathanklein jkle.in/craft Our Example Estimated Development Time: 4 Weeks Estimated
Development Cost: $8000 Client Is Paying: $12,000 54

@jonathanklein jkle.in/craft Our Example Estimated Development Time: 8 Weeks Estimated
Development Cost: $16,000 Client Is Paying: $12,000 55

@jonathanklein jkle.in/craft Software Engineer “Okay, we’ll cut some corners” 57

@jonathanklein jkle.in/craft Hyperbolic Discounting Given two similar rewards, humans show
a preference for one that arrives sooner rather than later. Humans are said to discount the value of the later reward, by a factor that increases with the length of the delay. 58

a preference for one that arrives sooner rather than later. Humans are said to discount the value of the later reward, by a factor that increases with the length of the delay. Solution 58

a preference for one that arrives sooner rather than later. Humans are said to discount the value of the later reward, by a factor that increases with the length of the delay. Solution Deliberately stop and do an estimation exercise 58

a preference for one that arrives sooner rather than later. Humans are said to discount the value of the later reward, by a factor that increases with the length of the delay. Solution Deliberately stop and do an estimation exercise Consider implications for future developers 58

@jonathanklein jkle.in/craft The Stanford Marshmallow Experiment 59

jkle.in/craft @jonathanklein Experimental Outcomes Researchers found that children who were
able to wait longer for the preferred rewards tended to have better life outcomes: • SAT Scores • Educational attainment • Body Mass Index (BMI) 63

@jonathanklein jkle.in/craft CEO “Who do we need to ﬁre?” 65

@jonathanklein jkle.in/craft Fundamental Attribution Error People's tendency to place an
undue emphasis on internal characteristics to explain someone else's behavior in a given situation, rather than considering external factors. 66

@jonathanklein jkle.in/craft Fundamental Attribution Error People's tendency to place an
undue emphasis on internal characteristics to explain someone else's behavior in a given situation, rather than considering external factors. Solution Blameless postmortems 67

@jonathanklein jkle.in/craft Everyone had good intentions 68

Our intuition is imperfect 69

Part 2: Real Studies

@jonathanklein jkle.in/craft A Quiz 72

@jonathanklein jkle.in/craft A Quiz A ball and a bat together
cost $1.10. The bat costs a dollar more than the ball. 72

@jonathanklein jkle.in/craft A Quiz A ball and a bat together
cost $1.10. The bat costs a dollar more than the ball. How much does the ball cost? 72

@jonathanklein jkle.in/craft A Quiz 73

@jonathanklein jkle.in/craft A Quiz Most people say $0.10 73

@jonathanklein jkle.in/craft A Quiz Most people say $0.10 The answer
is actually $0.05 73

@jonathanklein jkle.in/craft We Can’t Always Trust Our Intuition 74

@jonathanklein jkle.in/craft Dinner Plates Soup/Salad bowls Dessert plates Cups Saucers
8, all good condition 8, all good condition 8, all good condition 8, 2 broken 8, 7 broken 8, all good condition 8, all good condition 8, all good condition None None 75 Set A: 40 Pieces Set B: 24 Pieces

jkle.in/craft @jonathanklein Three Groups in the Experiment 1. Shown both
sets together 2. Shown only set A 3. Shown only set B 77

@jonathanklein jkle.in/craft Group 1 - Joint Evaluation: Set A: $32
Set B: $30 78

@jonathanklein jkle.in/craft Groups 2 and 3 - Single Evaluation: 79

@jonathanklein jkle.in/craft Groups 2 and 3 - Single Evaluation: Set
A: $23 Set B: $33 79

Manipulation

jkle.in/craft @jonathanklein 82

jkle.in/craft @jonathanklein Subscriptions to The Economist 82

jkle.in/craft @jonathanklein Subscriptions to The Economist 1. A web-only subscription
for $59 82

for $59 2. A print-only subscription for $125 82

for $59 2. A print-only subscription for $125 3. A web + print subscription for $125 82

for $59 2. A print-only subscription for $125 3. A web + print subscription for $125 83 Useless

for $59 2. A print-only subscription for $125 3. A web + print subscription for $125 84 16% 0% 84%

for $59 2. A print-only subscription for $125 3. A web + print subscription for $125 85

for $59 2. A print-only subscription for $125 3. A web + print subscription for $125 86 16% 84% 68% 32%

@jonathanklein jkle.in/craft Removing the middle option costs $3400+ per 100
subscribers 87

@jonathanklein jkle.in/craft Based on 2013 circulation, this could cost The
Economist $53M 88

@jonathanklein jkle.in/craft What’s the best way to sell a $3,000
suit? 89

@jonathanklein jkle.in/craft What’s the best way to sell a $3,000
suit? Put it next to a $10,000 suit 89

@jonathanklein jkle.in/craft Revisiting the Sunk Cost Fallacy 95

@jonathanklein jkle.in/craft It Preys on Loss Aversion 96

@jonathanklein jkle.in/craft Another Experiment 97 $0.01 $0.15

@jonathanklein jkle.in/craft Another Experiment 98 Most people choose this $0.01
$0.15

@jonathanklein jkle.in/craft Lower the Price 99 Free! $0.14

@jonathanklein jkle.in/craft Lower the Price 100 Most people choose this
Free! $0.14

@jonathanklein jkle.in/craft We need a framework for decision making 102

Bayesian Reasoning

@jonathanklein jkle.in/craft 104 Bayes' Theorem

@jonathanklein jkle.in/craft 1. Anchor your judgement of the probability of
an outcome on a plausible base rate 105

@jonathanklein jkle.in/craft 2. Question the importance and relevancy of your
evidence 106

@jonathanklein jkle.in/craft “What is the base rate for estimation accuracy?”
108

@jonathanklein jkle.in/craft “Why is this situation different?” 109

@jonathanklein jkle.in/craft To make rational decisions, you must understand the
base rate 115

@jonathanklein jkle.in/craft You must also objectively evaluate the evidence for
departing from the base rate 116

Two Principles: Dramatically better decisions 117

@jonathanklein jkle.in/craft 118 jkle.in/craft

@jonathanklein jkle.in/craft Thanks! jkle.in/craft @jonathanklein [email protected] 119

Cognitive Biases in Engineering Organizations - Craft Conf

Cognitive Biases in Engineering Organizations - Craft Conf

More Decks by Jonathan Klein

Other Decks in Technology

Featured

Transcript