Citation
The Effect of Education on Framing Effects and Sunk-Cost Errors

Material Information

Title:
The Effect of Education on Framing Effects and Sunk-Cost Errors
Creator:
Tallman, Lynn
LeBoeuf, Robyn ( Mentor )
Place of Publication:
Gainesville, Fla.
Publisher:
University of Florida
Publication Date:
Language:
English

Subjects

Genre:
serial ( sobekcm )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.

Downloads

This item has the following downloads:


Full Text




JOurn.31 or in.lnerr.3adu.3- Re--search

.. Oluinie , issue 2 - OctO:o er 21:11:14



The Effect of Education on Framing Effects and Sunk-Cost Errors

Lynn Tallman


ABSTRACT


This study explores the effect of education on framing and sunk-cost errors. Subjects were presented with a list

of decision scenarios both before and after an informative article. Reading information on how the framing of

a problem could bias decisions reduced the effect framing had on related decision scenarios. The same result

was witnessed for sunk-cost effects; education decreased the influence of sunk costs on decision problems.



INTRODUCTION


Everyday, people face a variety of unavoidable decisions, ranging from routine to critical. Everyone would like

to make the best decision possible based on the expected outcome, but this isn't always the case. Often,

decisions are influenced by inconsequential factors like wording changes and irrelevant information.



The normative approach to decision making assumes a rational decision maker who has well-defined

preferences (Shafir & Tversky, 1995). However, decisions are sensitive to how choice options are described.

For example, the same lottery may seem more attractive, possibly generating more ticket sales, when one

highlights that a participant has a 10% chance of winning as opposed to when one highlights that there is a

90% chance of losing. Such a pattern of preferences is known as a framing effect. "Framing effects are said to

occur whenever alternative descriptions of what is essentially the same decision problem give rise to

predictably different choices" (LeBoeuf & Shafir, 2002).



People succumb to other biases in decision-making, such as sunk-cost effects. A sunk cost is a past investment

of money, effort, or time associated with a current decision (Larrick et. al., 1990). According to traditional

economic theory, decisions should be based only on the incremental costs and benefits that are expected to

arise from an option because sunk costs cannot be recovered. Considering sunk costs while evaluating

current options might cause a person to make an irrational decision; and yet, people often honor sunk costs,

perhaps because they do not want to feel that they have wasted the already-spent investment. For example,

a person might have pre-purchased a $10 movie ticket and would still attend the movie, even if a better event

came up at the same time, like a free concert, that he or she would enjoy so much more than the movie.



It might be possible to improve decisions by teaching people the common errors of decision making and how





their preferences are often swayed by factors like framing and sunk-cost effects. If a person is aware of the

tendency to be biased by irrelevant information or problem framing, he or she might recognize when to

apply normative rules correctly. Shafir and Tversky (1995) concluded that when confronted with the fact that

their choices violate description invariance, people typically wish to modify their behavior to conform to

rationality principles. Therefore, a normative intuition exists within people. However, to fully debias people, they

have to know when to apply a learned rule and when to generalize it for similar problems (Wilson & Brekke, 1994).



Framing effects and the sunk-cost effect are errors in decision making according to the rational theory of

choice. Previous studies to test methods of preventing violations of the normative principle have yielded

mixed results. For example, experts are no more immune to these effects than are novices (Arkes & Blumer,

1985), but there is some evidence that encouraging subjects to think more about decisions decreases the

likelihood of exhibiting certain non-normative decision patterns (Sieck & Yates, 1997).



It seems that understanding the decision problem and possibly devoting more thought to a decision can reduce

the incidence of framing and sunk-cost effects. But because many studies have yielded mixed results,

exploring another, more direct, method of debiasing would shed light on the situation. In particular, using

education to explain the impact that inconsequential factors tend to have on decision-making could help

subjects understand the situation better and lead them to make rational decisions. The purpose of the present

study was to determine the effect education has on framing and sunk-cost in decision making.



METHOD


Subjects


Participants were one hundred and eighty-four University of Florida undergraduate students enrolled in either

the Introduction to Marketing or Quantitative Business Statistics course. For participating in the study, each

person was compensated with one point of extra credit.



Materials and Procedure


Each subject responded to four decision scenarios, then read a short educational article, followed by four

more decision scenarios. Four of the scenarios were designed to test for framing effects and four designed to test

for sunk-cost effects. The four framing scenarios and four sunk-cost scenarios were split into two pre-education

and two post-education each. Every scenario had a positive and a negative version; participants saw only one

version of each situation. For example, some subjects read this version of the decision problem, a positive frame,

and were asked to rate the taste, greasiness, quality, and lean-to-fat ratio on a 1-7 scale (Levin, 1987):


Suppose you are in the grocery store, and you find a package of ground beef that is labeled as 75% lean. How





would you rate that ground beef on each of the scales that follow?


Other subjects read this version of the decision problem, the negative frame:




Suppose you are in the grocery store, and you find a package of ground beef that is labeled as 25% fat. How

would you rate that ground beef on each of the scales that follow?



For sunk cost scenarios, one version had no sunk cost present in the decision problem while the other

version included a sunk cost in the information. For example, some subjects read (Gourville & Soman, 1998):




Your neighbor's cousin had a $40 ticket to a concert this weekend, but he can no longer attend. The neighbor

gives you the ticket and you are anticipating going to the concert. On the morning of the concert, there

are continuous torrential downpours and very dangerous road conditions due to heavy rain and widespread

flooding. Would you venture through the dangerous conditions and floods and still go to the concert?



Other subjects read this version:




You purchase a $40 ticket to a concert this weekend and you are anticipating going to the concert. On the morning

of the concert, there are continuous torrential downpours and very dangerous road conditions due to heavy rain

and widespread flooding. Would you venture through the dangerous conditions and floods and still go to the concert?



Thus, for each scenario, we could compare the responses of subjects who saw one version to those who saw

the second version of the problem. Furthermore, since each scenario appeared before education for some

participants and after education for others, we were able to determine whether the framing or sunk-cost effect

was reduced after education.



Three articles were used to test the effects of education as a debiasing method. Two were experimental, one

about framing effects and the second about sunk costs. Both of the experimental articles introduced the concept

and how the effect could unknowingly affect decisions. These articles also gave a basic example of framing effects

or sunk costs and explained the answer according to the rational theory of choice. The third article was a control,

an unrelated article on Christmas shopping.



RESULTS


The answers for each version were broken down by question and the percentage of participants that gave






each response was calculated (Table 1). Before education, subjects were affected by framing for all problems.

Those effects reached statistical significance for the Asian Disease _2(1) = 15.89, p < 0.001, N = 88, Sure win/

loss, _2(1) = 10.33, p < 0.001, N = 96, and average score of the Meat Rating problems, t(94) = 3.87, p <

0.000. Post-education results show that framing effects were significantly reduced (Figures 1 & 2) after subjects

read framing education for the Sure win/loss, _2(1) = 2.33, p < 0.127, N = 30, and average Meat Rating scenarios,

t(28) = 0.85, p < 0.402. The Chance to Win Game (Figure 3) also showed similar results, _2(1) = 0.01, p < 0.907,

N = 28, but the effects were not significant pre-education, _2(1) = 2.33, p < 0.127, N = 88. The influence of

framing effects was also occasionally decreased by sunk cost education for Asian Disease, _2(1) = 3.05, p < 0.081,

N = 37, and average Meat Rating, t(28) = 1.32, p < 0.198, scenarios (Figures 4 & 2). Surprisingly, the control

post-education results show a reduced framing effect in the Asian Disease, _2(1) = 0.80, p < 0.372, N = 31,

and Chances to Win game, _2(1) = 0.03, p < 0.853, N = 31, but again, this scenario was not significant

pre-education.



Table 1

Number of Subjects and Percentage of Sure and Risk Decisions by Pre-Education, Post Framing and Sunk Education,

and Control Group for Each Decision Scenario


Decision Scenario
Asian Disease


Pre-Education

n Sure Risk


Post Framing
Education
n Sure Risk


Post Sunk Education

n Sure Risk


Post Control

n Sure Risk


Positive Frame


Negative Frame

Game to Win


One-Stage Frame

Two-Stage Frame


Sure Win / Loss


51 74.51% 25.49%

45 42.22% 57.78%


17 58.82% 41.18%

13 30.77% 69.23%


14 71.43% 28.57%

16 31.25% 68.75%


15 80.00% 20.00%

13 38.46% 61.54%


Meat Rating

Lean/Positive

Fat/Negative

Airplane


46 32.61% 67.39%

42 78.57% 21.43%


16 12.50% 87.50%

12 83.33% 16.67%


20 0.15 0.85

17 82.35% 17.65%


15 26.67% 73.33%

16 68.75% 31.25%


Computer System


No Sunk Cost

Sunk Cost


6.25%

50.00%


60.87% 39.13%

19.05% 80.95%




38.10% 61.90%

54.35% 45.65%


93.75%

50.00%




66.67%

68.75%


25.00%

52.94%


75.00%

47.06%




35.29%

60.00%


33.33%

31.25%


40.00%

25.00%




50.00%

46.67%


60.00%

75.00%




50.00%

53.33%


64.71%

40.00%


45 3.3

51 4.3


13 4.19

17 4.56


No Sunk Cost

Sunk Cost


16 3.55

14 4.18


13 3.25

15 4.2


86.27% 13.73%

48.89% 51.11%


76.47%

61.54%


23.53%

38.46%


78.57%

75.00%


21.43%

25.00%


73.33%

61.54%


26.67%

38.46%






Concert Ticket


No Sunk Cost

Sunk Cost


60.00% 40.00%

62.75% 37.25%


46.15% 53.85%

23.53% 76.46%


31.25% 68.75%

64.29% 35.71%


23.08% 76.92%

53.33% 46.67%


Ski Trip


42 97.62% 2.38%



46 60.87% 39.13%


12 100.00% 0.00%



16 56.25% 43.75%


17 100.00% 0.00%



20 75.00% 25.00%


16 6.25% 0.00%

1.67%


15 5.00%


t Gann Frmet U Lou. Fnramv

foom% I mm-mL




Pro Ed~u P Fr rming Po Swk Edu Pos Cor*o
Edu



Figure 1. $300/$500 Sure Win/Loss.











Pft Edo Ps F raingr Poo Sutnk Edo pd CorArl
Edu


Figure 3. Chance to Win Game.


* Peti Fvie I Negs~e Frime

Soo




Pre "du Pos Frarmn1 Po Sok Edu Pos COnto
E4d3



Figure 2. Meat Rating Problem.












Edu


Figure 4. Asian Disease Problem.


Pre-education, results show sunk-cost effects for all scenarios (Table 1) except the Concert Ticket problem, _2(1)

= 0.08, p < 0.783, N = 96. Sunk-cost effects reached statistical significance in the Airplane Investment, _2(1)

= 18.71, p < 0.000, N = 88, Computer System Purchase, _2(1) = 15.55, p < 0.000, N = 96 and Ski Trip

scenarios, _2(1) = 17.51, p < 0.000, N = 88. Post-education results reveal that sunk costs effects were reduced

in the Computer System Purchase scenario (Figure 5). The reduction was greatest after sunk cost education, _2(1)

= 0.05, p < 0.818, N = 30, but framing education also reduced effects at _2(1) = 0.78, p < 0.376, N = 30.

Reading the control article actually reduced sunk costs effects post-education for the Computer System, _2

(1) = .044, p < 0.505, N = 28, and Ski Trip scenarios, _2(1) = 1.30, p < 0.254, N = 31.


No Sunk Cost


Sunk Cost


















Figure 5. Computer System Purchase.



DISCUSSION


According to the normative approach to decision making, a rational strategy is one that will produce results that

the decision maker desires (Pious, 1993). However, a lack of clarity or knowledge allows people to be led astray

by problem cues such as framing or irrelevant information such as sunk costs. The success of attempts to

increase people's awareness of bias and influence decision making depends in part on the extent to which

researchers can convince subjects that their judgments are, in fact, open to bias while motivating them to correct

it (Wilson & Brekke, 1994). In this experiment, framing and sunk-cost education has proven to be a method

of convincing subjects that their judgments are biased for some decision scenarios.



The framing effects that were most reduced by education all had something in common: it was easy to consider

an opposite frame and obtain an unbiased view of the problem. The Meat Rating scenario, featured above, is a

prime example. Due to the nature of the question, it should be rather simple to convert from the positive frame

to the negative, or vice versa, and then infer that the meat, for example, is not just X% fat but also (100-X)%

lean (LeBoeuf & Shafir, 2003). This is also true for the $300/$500 Sure win/loss scenario (Tversky &

Kahneman, 1986). Some subjects read the positively framed version:




Imagine you are $300 richer than you are today. What would you prefer?




A. Winning $100 for sure

B. A 50 percent chance at winning $200, and a 50 percent chance at winning nothing



Other subjects read the negatively framed version:


Imagine you are $500 richer than you are today. What would you prefer?


*No SunkCd Surnk Coad
I 10D% j _ _ _ .._ ^^^



J% J J J
S Pro EJ PoWFrinirng P"SwunhkE PstCnwHl
Edu






A. Losing $100 for sure

B. A 50 percent chance at losing $200, and 50 percent chance at losing nothing



In this scenario, it is easy for a person to calculate the expected utility of each decision choice to obtain the

exact dollar gain or loss. Framing education made participants aware of the frame and the need to calculate the

pay out of each choice to make a rational decision.



The sunk-cost scenarios used in the experiment provide enough information to draw a few conclusions.

This experiment supports the theory that "people who have learned the principle of sunk costs are less likely to

make certain kinds of errors in reasoning than those who have not" (Wilson & Brekke, 1994). Framing effects

and sunk costs had a much more significant impact on participants before sunk cost education. So, it can be

said that, after reading sunk cost education, subjects are less likely to make certain kinds of errors in reasoning.

This was especially true for the Computer System Purchase scenario (Arkes & Blumer, 1985).



Some subjects saw this version, containing no sunk cost:




As the owner of a small business, you must choose whether to modernize your operation by spending $200,000 on

a new computer system or on a fleet of new delivery trucks. You choose to buy the trucks, which can deliver

your products twice as fast as your old trucks at about the same operating cost as the old trucks. One week

after your purchase of the new trucks, one of your competitors goes bankrupt. To get some cash in a hurry, he

offers to sell you his top-of-the-line computer system for $10,000. There is no question that his system is better

than your current system: his system works 50% faster than your current system at about one-half the

operating cost. You will not be able to sell your current computer system to raise this money, since it was

built specifically for your needs and cannot be modified. However, you do have $10,000 in savings. Should you

buy the top-of-the-line computer system from your bankrupt competitor?



Other subjects read the version with a sunk cost included:




As the owner of a small business, you must choose whether to modernize your operation by spending $200,000 on

a new computer system or on a fleet of new delivery trucks. You choose to buy the computer system, which

works twice as fast as your old computer system at about the same operating cost as the old system. One week

after your purchase of the new system, one of your competitors goes bankrupt. To get some cash in a hurry,

he offers to sell you his top-of-the-line computer system for $10,000. There is no question that his system is

better than your new system: his system works 50% faster than your new system at about one-half the

operating cost. You will not be able to sell your new computer system to raise this money, since it was

built specifically for your needs and cannot be modified. However, you do have $10,000 in savings. Should you





buy the top-of-the-line computer system from your bankrupt competitor?


After reading the sunk cost education article, it is no surprise that subjects would now understand the possibility

of bias in this problem and that sunk-cost effects would be reduced post-education. Education helped

participants withhold their desire not to appear wasteful and make the investment in the computer system.



In conclusion, education showed some effect on debiasing framing effects and sunk-cost errors. Further

studies should be conducted to reaffirm the results, using more subjects and a broader range of scenarios.






REFERENCES



1. Arkes, H.R., & Blumer, C. (1985). The Psychology of Sunk Cost. Organizational Behavior and Human

Decision Processes, 35, 124-140.

2. Gourville, J.T., & Soman, D. (1998). Payment Depreciation: The Behavioral Effects of Temporally

Separating Payments from Consumption. The Journal of Consumer Research, 25 (2), 160-174.

3. Larrick, R.P., Morgan, J.N., & Nisbett, R.E. (1990). Teaching the Use of Cost-Benefit Reasoning in Everyday

Life. Psychological Science, 1 (6), 362-370.

4. LeBoeuf, R., & Shafir, E. (2003). Decision Making. Manuscript in preparation, University of Florida.

5. LeBoeuf, R., & Shafir, E. (2002). Deep Thoughts and Shallow Frames: On the Susceptibility to Framing

Effects. Journal of Behavioral Decision Making, 15, 1-16.

6. Levin, I.P. (1987). Associative Effects of Information Framing. Bulletin of the Psychonomic Society, 25 (2), 85-86.

7. Pious, S. (1993). The Psychology of Judgment and Decision Making. Philadelphia, PA: Temple University Press.

8. Shafir, E., & LeBoeuf, R. (2002). Rationality. Annual Review of Psychology, 53, 491-517.

9. Shafir, E., & Tversky, A. (1995). Decision Making. In E. E. Smith 7 D.N. Ohserson (Eds.), Thinking: An Invitation

to Cognitive Science, 3, 77-100. Cambridge, MA: The MIT Press.

10. Sieck, W., & Yates, J.F. (1997). Exposition Effects on Decision Making: Choice and Confidence in

Choice. Organizational Behavior and Human Decision Processes, 70, 207-219.

11. Tversky, A., & Kahneman, D. (1986). Rational Choice and the Framing of Decisions. Journal of Business, 59 (4),

251-278.

12. Wilson, T.D., & Brekke, N. (1994). Mental Contamination and Mental Correction: Unwanted Influences on

Judgment and Evaluations. Psychological Bulletin, 116 (1), 117-142.





--top--



Back to the Journal of Undergraduate Research


College of Liberal Arts and Sciences I University Scholars Program I University of Florida |


� University of Florida, Gainesville, FL 32611; (352) 846-2032.


SJ UNIVERSITY of
UK FLORIDA
The � otrdaIfsh. tofor Fir Go w r NahioT