reflections

1

Choice Context

BUS143: Judgment and Decision Making

Ye Li

Repeating themes in this class

• People’s evaluations tied to the local, rather than global context. For example: – (Topic 1) We take choices as given (concreteness

principle), and evaluate outcomes relative to reference points (prospect theory)

– (Topic 4a) We form narrow, “topical” accounts rather than comprehensive mental accounts

– (Topic 5) We exhibit myopia in intertemporal choices

• Why? – In most cases, people find relative evaluation easier than absolute

evaluation

2

Choice in context

• Given a set of alternatives, how do people select a preferred option?

• How about buying a new smartphone? – What was your process like? – How would you describe it?

• How about what school to attend (or apply to)? • How about what to eat for breakfast?

Making choices: What do Econs do?

• A value-maximizing decision-maker would… – Take stock of goals (i.e., knows exactly what he or

she wants) – Explore ALL alternatives – Evaluate how well each alternative addresses their

goals – Choose alternative that has greatest total utility

3

Value Maximization: Prescriptions from Intro Economics

• Use a decision matrix (step 1 of 3) – First, identify and set priorities among objectives  Assign attribute weights (i.e., importance)

Attribute Importance 4 3 1 2 2 5 3 2 3

Attribute Weight 16% 12% 4% 8% 8% 20% 12% 8% 12%

Option Price Size Weight Display Camera Software Storage Processor Battery

Value Maximization: Prescriptions from Intro Economics

• Use a decision matrix (step 2 of 3) – Second, determine how alternatives measure up  Assign attribute values  Scale each dimension from 0 (worst) to 100 (best)

Option Price Size Weight Display Camera Software Storage Processor Battery

iPhone X 0 40 0 75 100 100 0 100 100

iPhone 8 67 95 80 0 50 100 0 95 50

iPhone 7 78 100 100 0 0 100 75 0 0

Galaxy S8 100 0 60 50 40 0 100 50 45

Option Price Size Weight Display (" PPI type) Camera Software Storage Processor Battery

iPhone X $1085 5.65×2.79×0.30 6.14oz 5.8" 458 OLED 12 dual/7 iOS 11 64gb A11 ~10:35

iPhone 8 $760 5.45×2.65×0.29 5.22oz 4.7" 326 IPS LCD 12/7 iOS 11 64gb A11 8:37

iPhone 7 $705 5.44×2.64×0.28 4.87oz 4.7" 326 IPS LCD 12/8 iOS 11 128gb A10 7:46

Galaxy S8 $600 5.85×2.68×0.31 5.36oz 5.8" 570 SAMOLED 12/8 Android 64gb+ Snap. 835 8:22

4

• Use a decision matrix (step 3 of 3) – Calculate utility for each option  Multiply values by weights, sum across options

Attribute Weight 16% 12% 4% 8% 8% 20% 12% 8% 12%

iPhone X 58.8 0 40 0 75 100 100 0 100 100

iPhone 8 62.9 67 95 80 0 50 100 0 95 50

iPhone 7 57.5 78 100 100 0 0 100 75 0 0

Galaxy S8 51.0 100 0 60 100 40 0 100 50 45

Value Maximization: Prescriptions from Intro Economics

Why not use this method for most choices you encounter in life?

Making Choices: What Humans actually do

• Humans use shortcuts – People often make “reason-based” choices (more

details later) – Screening (removing options) – Relative rank matters (not absolute goodness)

• Implications… – For modeling people’s choice behavior – For product positioning  Also, biological, medical, legal, and political implications

5

X4=taste

X3=calories

Homo economicus: How Econs maximize value

X2=sugar

X1=caffeine

“Conjoint analysis” (a major marketing tool) is based on assumption that utility of option is sum of component utilities (“purely additive model”)

Economic Modeling of Choice I: 1950-1970s

Coke 60%

Pepsi 40%

Coke 48%

Pepsi 32%

TALLP (ex-Pepsi)

8%

TALLP (ex-Coke)

12%

Assumption: Proportionality (“constant ratio rule”) New offering will take in proportion to original shares.

TALLP

Suppose that, when added, TALLP takes 20% share

6

Choice Modeling II: Similarity Hypothesis

Coke 60%

Pepsi 40%

TALLP

Coke 55%

Pepsi 25%

TALLP (ex-Pepsi)

17%

TALLP (ex-Coke)

3%

Assumption: Similarity New offering will take more share from those that are similar

(i.e., similar goods swap out for each other in the market)

Again, suppose that TALLP takes 20% share

Choice Modeling III: Regularity Assumption

x

z

y

A B

Pr(x;A) = ? Pr(x;B) = ?

Pr(x;A) ≤ Pr(x;B)

The entry of an additional alternative will either reduce the share of existing alternatives or leave them unchanged

7

High-Stakes Violation of Regularity Redelmeier & Shafir 1995

• Scenario presented to neurosurgeons: Who has priority for surgery? – Two options  Woman: mid fifties  Man: early seventies

– Three option  Woman: mid fifties  Man: early seventies  Woman: mid fifties

• Why do these surgeons violate regularity?

C

T

Quality

Price

70

50

$1.80$2.60

D

C = competitor

D = decoy

Violating choice principles: The Attraction Effect Huber, Payne, & Puto 1982

T = target

8

B

A Similarity

Distance (in miles)

80

70

3050

C

The Attraction Effect in Dating

60

35

Context Effect 2: Decoys without dominance?

C

T

Quality

Price

70

50

$1.80$2.60

D

Efficient Frontier

“Compromise Effect” - Not just similarity effect - Not necessarily a relatively inferior alternative

9

B

A Similarity

Distance (in miles)

80

70

50

C

The Compromise Effect in Dating

90

75 30

Why? Extremeness Aversion (harder to defend)

A

Probability of Repair

40

16

B

C

32

24

8

9% 7% 5% 3% 1%

Shift from B to A = loss of reliability

Shift from B to A = gain of functionality

Shift from B to C = loss of functionality

Shift from B to C = gain of reliability

Number of functions

10

Reason-Based Choice Shafir, Simonson & Tversky 1993

• Basic idea: Individuals construct reasons to resolve conflict and justify their choice – “Choice is a search for a unique principle that covers the

decision at hand and is not dominated by another more powerful principle.” (Prelec & Hernstein 1991)

– Reason-based choice seems more compelling than a tradeoff-based choice

• Why is making tradeoffs difficult? – Conflicting objectives (and loss aversion!)  Tradeoffs create losses on some attributes

– Optimizing/maximizing (pick the best) versus Satisficing (pick something that is ‘good enough’)

Reason-based Choice II Shafir, Simonson & Tversky 1993

• Reason-based choice is NOT normative because: – “More important” attributes get too much weight in reason-

based choice – Reasons are frame-dependent (see next slide)

• Reason-based choice occurs more often: – In complicated situations (lots of information; many

alternatives) – When value-based approaches are hard to defend

 E.g., tradeoffs about human lives, safety, etc.

Would you expect context effects to be exacerbated or diminished in organizational decisions?

• Reason-based choice increases with accountability!

11

Which to choose? Shafir 1993

Imagine that you serve on the jury (one of 12 jurors) of an only- child sole-custody case following a relatively messy divorce. The facts of the case are complicated by several ambiguous economic, social, and emotional considerations, and you must decide on the basis of the following few observations:

To which parent would you award sole custody of the child? • Parent A

– average income, average health, average working hours, reasonable rapport with the child, relatively stable social life

• Parent B – above-average income, very close relationship with the child,

extremely active social life, lots of work-related travel, minor health problems

reject?

Give people a reason to choose you… Iyengar & Lepper 2000

• Field experiment • Two tasting booths

A. Few options: 6 jams

B. Many options: 24 jams

• Difficult to make choice  No choice or choice deferral

• Your examples?

12

Disjunction Effect Tversky & Shafir 1992

Imagine that you have just taken a very tough final exam. It is the end of Winter quarter, you feel tired and run-down, and you are not sure that you passed the exam (or class). In case you failed, you have to take the class again in spring quarter—after Spring Break. You now have an opportunity to buy a very attractive 5- day vacation package in Hawaii at an exceptionally low price. The special offer expires tomorrow, while the exam grade will not be available until the day that. Would you…?

A. Buy the vacation package. B. Not buy the vacation package. C. Pay a $5 non-refundable fee in order to retain the rights to

buy the vacation package at the same exceptional price the day after tomorrow—after you find out whether or not you passed the exam.

Context matters for evaluability Hsee et al 1999

• Joint evaluation: Simultaneous consideration of two or more options

• Separate evaluation: Consideration of each option in isolation

• Example: How happy would you be with this legal settlement? A. You get paid $500 and other person gets paid $500 B. You get paid $600 and other person gets paid $800

• Evaluability of a continuous attribute depends on knowledge of average, best and worst values – Reflects the desirability of an attribute value in a given

decision context

13

Evaluability Hypothesis Hsee et al 1999

• When an option is judged in isolation, the judgment is influenced more by the attributes that are easier to evaluate in isolation – E.g., salary, beauty, height… What else? – Give me examples of dimensions that are or are

NOT evaluable. (e.g., diamond 4C’s)

• In joint evaluation, each option serves as the most salient reference for evaluating the other options  Changes evaluability – Can shift what attributes are important!

Joint evaluation increases evaluability Hsee 1996

Assume you are a music major and are looking for a music dictionary in a used book store and planned to spend between $10 and $50.

$22.05 

$20.20 

$18.23 

$25.25 

 $15.00

 $17.00

 $19.00

 $21.00

 $23.00

 $25.00

 $27.00

Separate Evaluation Joint Evaluation

Your Data

Dictionary A

Dictionary B

14

More Joint vs. Separate Reversals Hsee 1998

Imagine that you are shopping for a dinnerware set. There is a clearance sale in a local store where dinnerware usually sells for between $40-$80 a set. Suppose that you have found these two sets (this set) on clearance. They are made by a reputable manufacturer and are white and simple.

$35.33 

$40.13  $41.13 

$37.33 

 $25.00

 $30.00

 $35.00

 $40.00

 $45.00

Separate Evaluation Joint Evaluation

Your Data Set A Set BSet A includes  40 pcs

Set B includes  24 pcs

Dinner plates: 8, all in good  condition

8, all in good  condition

Soup/salad  bowls:

8, all in good  condition

8, all in good  condition

Dessert plates: 8, all in good  condition

8, all in good  condition

Cups: 8, 2 of them  are broken

Saucers: 8, 4 of them  are broken

Which evaluation mode is better?

• Of course, neither is always better, so qualify your answer with when.

• Most researchers would argue for joint evaluation. Why?

• When and why might separate evaluation be better?

15

Summary

• Choice depends on context – Marketers should consider how to control the choice

environment, e.g. via product characteristics. – Decisions in separate and joint evaluations may differ!

• People hate making tradeoffs (loss aversion) and generate reasons to choose one alternative over the other(s)

• This leads to violating basic choice principles – The Attraction Effect – The Compromise Effect

Ye’s Keys

14. Everything is relative. Context impacts choice and evaluability.

15. Change the context  change the choice: Maximize your position by identifying extremes and beating the heck out of the weak competitor.

16. Easier to make choices we can explain to others (and ourselves). So give people a reason to choose your product!

17. You only live one life, so some attributes are hard to evaluate (e.g,. diamonds). Make sure you are maximizing the ‘right’ attributes in choices you make in joint evaluation.