Management of Technology Projects: Why Projects Fail and How to Plan For Disputes

In its seminal 1995 study, the Standish Group reported the dismal success rate of
large scale information technology (“IT”) projects. The Standish Group’s research showed that
31.1% of projects were cancelled without completion, while only 16.2% of software projects
were completed on-time and on-budget. Large companies reported even worse results as only
9% of their projects came in on-time and on-budget. Further, when projects were completed,
they often fell far short of their expectations: projects completed by the largest American
companies had only about 42% of their originally-proposed features and functions.
Although more recent studies by the Standish Group suggest improved success
records, the record hardly encourages optimism. In 1995, fully 84 percent of all projects were
found to have failed entirely or to live up to objectives. By 2000, this failure rate had
“improved” to 72 percent.

The Standish Group reports are hardly alone in reporting the dismal record of IT
project success. In 2005, KPMG International surveyed 600 organizations across 22 countries
and found that nearly half of the organizations reported at least one project failure in the past
year, and 86% of respondents reported failure to achieve 25% their targeted benefits.
Given this poor track record over an extended period of time, why have IT
suppliers seemingly failed to learn from past mistakes and why have their customers failed to
2 © Gary L. Kaplan, Esq. 2008, All Rights Reserved.

Page 2
demand greater accountability? As discussed below, the answer may lie in inherent human
characteristics that allow seemingly irrational conduct to persist in even the most sophisticated
business environments. Scientific research over the past 20 year has largely debunked the notion
that Adam Smith’s “invisible hand” assures the efficient economic transactions. Instead,
individual economic decisions commonly follow predictable, but seemingly irrational, patterns
that allow for inefficient market conditions to persist.
The good news, however, is that when an understanding of the factors that
promote irrational expectations is coupled with an understanding of circumstances that
commonly cause IT projects to fail, it is possible to develop systems for tracking project
performance and managing disputes that dramatically improve prospects for success.
The Persistence of Doomed IT Projects
Social scientists have identified, tested and explained a variety of human character
traits, or biases, that are particularly relevant to the planning and management of complex
technology (or even construction) projects.
At the outset, it is perhaps more surprising that complex IT projects ever succeed
than that they have such a high failure rate. Although there is likely a single path to project
success, there are multiple (and easy) paths to project failure. Common sources of IT project
failure include:
• Incomplete project specification
• Poor project management
• Overselling
• Personnel changes
• Technology changes

Page 3
• Poor communications between technology and business personnel
• Failure to coordinate technology with business practice/culture
• Data migration failures
• Poor software design
• Infrastructure incompatibilities
• Regulatory change
Reduced to its simplest terms, most IT project fail sufficiently to address or
account for (i) human and non-technology factors that may affect project success or failure, and
(ii) risks related to changing circumstances. In a long term IT project, it may be impossible to
predict which circumstances affecting project success will change, but it is almost certain that
important circumstances will, in fact, change.
Against this backdrop, this paper discusses the persistent failure of IT
professionals and their clients to plan for changing circumstances that may undermine project
success and suggests strategies for promoting accountability and managing inevitable disputes.

A. The Persistence of Unrealistic IT Project Plans and Agreement

Human nature encourages largely unrealistic project plans and related agreements
governing complex IT projects. Such traits include the “overconfidence” bias and illusions of
control (i.e., the tendency to overstate one’s ability to affect outcomes); the “over-optimism” bias
(i.e., the tendency to understate the likelihood of bad outcomes); the “self-serving” bias (i.e., the
tendency to evaluate facts in the light most favorable to oneself ); the “confirmation” bias (i.e.,
the tendency to filter and interpret new information in a manner that confirms one’s own
preconceptions or initial view); and the “availability” bias (i.e., the tendency to overstate the
importance of available information). These problems are magnified, moreover, because of the

Page 4
conflicting interest’s of participants in complex technology projects, which discourages
meaningful accountability.
While such biases may not apply in all IT project failures, taken together they
help explain the unfortunate persistence of overoptimistic project plans and agreements.

1. Overconfidence, Undue Optimism and Illusions of Control

Planning for IT projects may suffer from an especially pervasive characteristic of
human nature--self-deceptive overconfidence. If asked about the likelihood of project success,
the representative of an IT supplier may honestly believe that his company will have the skill and
experience to complete the project on time, on budget, and in accordance with its specifications.
Such expectations, however, are likely to be severely skewed as a result of both human nature
and self interest. For example, one study found that fully 94 percent of university professors
believe that they do a better job than their colleagues.3 Similarly, most people think that they are
more intelligent and fair-minded than average.4
Overconfidence may be closely related to irrational forces affecting decisionmaking:
individuals often believe that they have more control over outcomes than they do.
Perhaps the best example of this phenomenon is the New Jersey lottery. The lottery was a failure
when ticket purchasers were given computer-selected numbers; it became a huge success when it
was revised to permit purchasers to select their own numbers—even though the odds of winning
were the same in either case.
Similarly, attorneys and their clients—like people in other circumstances—are
prone to vastly overestimate the likelihood of positive outcomes and underestimate the likelihood
3 (Cross, 1999).
4 (Birke & Fox, 1999, p. 7).

Page 5
of negative ones. In an early study, college students rated themselves much less likely than their
peers to suffer from a drinking problem, have a heart attack, be fired from a job or divorce a few
years after getting married.5 Similarly, one review of the academic literature revealed the
following:
• Second-year M.B.A. students overestimated the number of job
offers they would receive and their starting salaries.
• Students overestimated the scores they would achieve on exams.
• Almost all the newlyweds in a U.S. study expected their marriages
to last a lifetime, even while aware of the divorce statistics.
• Professional financial analysts consistently overestimated
corporate earnings.
• Most smokers believe they are less at risk of developing smokingrelated
diseases than others who smoke.6
Similar findings of overoptimism were found in studies of trial lawyers. The authors of one
study, for example, concluded that “lawyers tended to be overconfident in general, but especially
so in cases in which they initially made highly confident predictions.”7

2. Self-Serving Bias: Confusing Fairness with Self-Interest

“However top-lofty and idealistic a man may be, he can always rationalize his right to
earn money.” –Raymond Chandler
Coupled with intrinsic over-confidence, and IT suppliers self-interest will
promote unrealistic (however honestly held) projections of project costs and benefits. Research
has shown that when a problem affects a party’s self-interest, that party’s analysis is likely to be
skewed to a far greater extent than he or she is likely to realize or admit.
Two cited studies illustrate how one’s team or position affects one’s perspective.

5 (Weinstein, 1980).
6 (Armor & Taylor, 2002).
7 (Loftus & Wagenaar, 1988).

Page 6
In a 1954 study,8 students from Princeton and Dartmouth were asked to review a film of a
football game between the two schools and to count the number of penalties by each side. The
Princeton students found that the Dartmouth team committed twice as many flagrant penalties
and three times as many mild penalties as the Princeton team. On the other hand, the Dartmouth
students found that the two teams committed an approximately equal number of penalties. The
study concluded that it was as though the two groups “saw a different game.”
Likewise, the “self-serving” bias was clearly demonstrated in a 1997 study by
Professors George Lowenstein and Linda Babcock that considered how a group of “plaintiffs”
and “defendants” evaluated a single hypothetical case.9 Specifically, Lowenstein and Babcock
asked college students who were divided into plaintiffs and defendants to evaluate a tort case in
which an injured motorcyclist sued the driver of the car that collided with him for $100,000. All
students were given the same twenty-seven-page case file and were then asked to predict how a
judge would rule and the amount that a neutral third party would consider to be a fair settlement.
The students were then given thirty minutes to negotiate a settlement and financial incentives to
resolve the matter efficiently. (Specifically, the students, who were paid for the study, would be
allocated payments based on the results of their settlement or the “court’s” ruling. In the absence
of a settlement, both parties would incur “court costs” that would reduce their payments.)
In the initial version of the study, the students were advised of their roles before
reviewing the case file. The parties’ self-serving biases were demonstrated in that the students
assigned to the plaintiff’s role predicted a judicial ruling that averaged $14,527 higher than

8 (Hastorf & Cantril, 1954).
9 (Babcock & Loewenstein, 1997).

Page 7
students assigned to the defendant’s role. Likewise, the student “plaintiffs” assessed the fair
settlement value of the case $17,709 higher than student “defendants.”
Next, Lowenstein and Babcock demonstrated that this bias was not simply a
product of the students’ assignment to a plaintiff or defendant role, but that it reflected their
skewed interpretations of the case-file: In a second round of the study, some students were
assigned to a control group that was instructed to review and assess the case file before their
designation as plaintiffs or defendants.
The second round confirmed that the students’ assigned role affected how they
first interpreted the case file: when they didn’t know which roles they would be assigned until
after they read the case and made assessments of the judge and fairness, the case “settled” 94
percent of the time. By contrast, when students knew what their roles would be before reviewing
the file, only 72 percent of the cases settled.
Finally, Lowenstein and Babcock sought to identify methods for counteracting the
self-serving bias. Simply by telling the students about the bias improved the accuracy with
which they predicted the valuations of the case to be made by the opposing party (but did not
induce them to make more accurate assessments themselves). If the students were told about the
bias and then asked to write an essay in support of their opponent’s position, the parties’ biases
actually increased—presumably because they led the students to reaffirm their commitments to
their own positions. On the other hand, Lowenstein and Babcock were able to counteract the
bias by advising the students of the bias and then instructing them to list the weaknesses in their
own case. Under this approach, the difference between the plaintiff’s and defendant’s average
assessment was reduced from $21,783 to only $4,674, and the rate of failed settlement
negotiation declined from 35 percent to 4 percent.

Page 8
Such research implies that initiating a dispute is in itself likely to affect how each
party considers the facts in dispute and to cause each side to interpret facts in the light most
favorable to its position. Although it may be possible for a third party to intervene and
counteract this bias, without a skilled mediator each party’s bias will probably just help fuel the
conflict. Worse, the self-serving bias does not stand alone in promoting skewed perceptions and
assessments of a dispute.

3. Poor Predictions in the Face of Incomplete Information

Research has shown that individuals faced with uncertainty or incomplete
information will selectively fill in details in ways that may promote conflict. Each party to an
agreement is likely to give more weight to the facts that they know or that easily come to mind
than to hold a rational view of what all the relevant facts suggest is merited. This trait has been
called the “availability heuristic.” (A “heuristic” is a mental shortcut used to evaluate complex
sets of facts.)
The forcefulness of the availability heuristic is shown by undue reliance on
anecdotal evidence or past instances of a particular event and outcome; i.e., reliance on the
erroneous belief that “because it happened before it will happen again.”
In research studies, the availability heuristic has been shown to apply not only to
incomplete information (such as when a party knows only half of the relevant events), but also to
more easily understood or accessible information. For example, psychologists Daniel Kahneman
and Amos Tversky, who wrote the seminal work on this topic,10 asked participants in a study to
estimate the number of words that begin with the letter “R” or “K” in comparison to words that
10 (Tversky & Kahneman, 1974).

Page 9
have the letter “R” or “K” in the third-letter position. Because it is far easier to recall words that
begin with “R” (rooster, roar, rusty, red) or “K” (key, kitchen, kite) than words having “R” or
“K” in the third-letter position (street, stream, ink, acknowledge), 66 percent concluded that
words that begin with “R” or “K” are more common. The reality is that words that have the
letter “R” or “K” in the third position are far more common than those beginning with those
letters.11
A more common example is to ask study participants whether more people die in
airplane crashes or car accidents. Because airplane crashes receive far more media attention,
many people erroneously believe that they are responsible for more deaths, although statistics
show that car crashes are responsible for far more fatalities.
In connection with a long term project, availability biases can create irrational
projection as to project benefits and costs. As noted above, at the outset of a project, each party
may give undue weight to readily accessible facts while ignoring the implications of the full
picture. In addition, parties are likely to underestimate the costs and stresses associated with the
project, because such negative consequences are remote from factors most commonly considered
at the outset.

4. No Way Out

Once a project goes off track, it can be difficult correct. The failure to address
project problems as they arise can lead to a cascade of additional and increasingly complex
difficulties. In the end, the project may have little in common with its original plans. Part of the
difficulty in addressing project problems in a timely manner stems from the cost of the project
11 (Kahneman & Tversky, 1982).

Page 10
itself: the further a project progresses, the more it costs, and the parties’ investment contributes
to intransigence and/or wishful thinking.
Psychologist and behavioral economists refer to this issue as the “sunk cost” bias
or “sunk cost” fallacy. In economic terms, “sunk costs” or “fixed costs” refer to investments that
have been made but cannot be directly recovered through future actions, regardless of the
success or failure of those actions. For example, if you buy a nonrefundable airline ticket to visit
your aunt Mabel, you will not be refunded the cost of the ticket if you cancel the visit.
Consequently, after the ticket has been paid for, its cost, according to economic theory, should be
irrelevant to the decision of whether or not to go ahead with the trip. If you have a fight with
Aunt Mabel and don’t want to go, the fact that you paid $1,000 for the ticket should have no
bearing on your decision, since you will not get the money back in any event.
Common experience (confirmed by research) shows, however, that people often
cannot bring themselves to ignore sunk costs when deciding future actions and will, in fact,
throw good money after bad to justify their prior mistakes. Part of the problem is analogous to
the self-interest bias: once a party has spent money on a case, the investment itself is likely to
taint his or her appraisal of the outcome. One study of this issue was made at a racetrack—
bettors on a horse race were asked to predict their horse’s chance of winning both before and
after having made a bet. Of the 141 people asked, 72 had just placed a $2 bet within the past
thirty seconds, and 69 were about to place a $2 bet in the next thirty seconds. On a scale of 1 to
7, with 7 referring to a sure winner, the people about to place a bet averaged a 3.48 rating of their
Page 11
chances for success (or a “fair chance of winning”), but people who had just finished betting
averaged a 4.81 rating (or “good chance of winning”).12
Further, people have a hard time admitting that they made a mistake. In another
study, 96 business students were divided into two groups: half were told that, as a manager, they
had made a bad R&D investment in an underperforming division of a company, while the other
half were told that the former manager of the division had made the bad R&D investment. In
both cases, the students were then asked to make a new $20 million investment (to be divided
between the underperforming division and another division). The students “responsible” for the
poor performance of the division were much more prone to invest in that division than the
nonresponsible students. On average, the “responsible” students chose to invest $12.97 million,
while the others only $9.43 million.13

B. Managing IT Project to Avoid and Contain Disputes

IT Disputes are entirely predictable. Just as preventative health care or routine
equipment maintenance can avoid far greater costs in the future, planning for disputes and
managing business relationships with the risk of disputes in mind can go a long way to
minimizing disastrous projects and costly legal battles.
Although there are a variety of methods for managing potential disputes, they
essentially rest on a single principle: project participants should strive to resolve disputes
promptly as a means of keeping small disputes small, rather than allowing them to fester and
transform themselves into corporate wars.
12 (Knox & Inkster, 1969). See also (Arkes & Blumer, 1985); (Arkes & Hutzel, The Role
of Probability of Success Estimates in the Sunk Cost Effect, 2000).
13 (Staw, 1976). See also (Whyte, 1986).

Page 12
The simple concept of containing disputes is difficult for two reasons. First, the
parties must have mechanisms in place to identify and quickly resolve problems as they arise.
Such mechanisms can, and should, be included in any complex transaction or engagement that
requires the parties to interact over a period of time—such as a construction project or software
development agreement. Second, the parties must actually deploy the mechanisms they have
agreed upon.
Perhaps surprisingly, the second requisite—use of available dispute
mechanisms—is where parties to a business arrangement often fall short. Once a project is
underway, the parties to a business arrangement often prefer to focus exclusively on moving
forward and “working things out” as they proceed with each other. Further, the individuals
implementing a project may be a completely different group of people than those who negotiated
the arrangement (and any dispute resolution mechanism). Indeed, project personnel may never
even look at the contract terms governing their relationship, preferring instead to relegate “legal
issues” to the back burner.
The problem with assuming that problems will work themselves out is that they
often do not. Further, issues or problems with complex projects seldom arise alone, but typically
arise in groups or a series. As a result, when a series of small issues are not addressed, the
project itself may take on a character that is quite distinct from the project originally
contemplated by the parties who negotiated the terms of the relationship. If the parties have not
addressed and modified their contract expectations along the way when the series of small issues
emerges as an unworkable mess, the cost of untangling the parties’ respective duties and
obligations can be extraordinary.

Page 13

1. Managing Potential Disputes with Effective Contracts

The starting point for managing potential risks is to negotiate contracts that both
clearly articulate the parties’ respective obligations and specify efficient procedures for resolving
disagreements as they arise.
In defining the parties’ obligations, an effective contract addresses the fact that
the parties may have multiple types of relationships with each other in a single transaction. For
example, when a business hires a software developer to create new inventory software, it may
need to rely on the vendor for a variety of services in addition to the software itself. The
business may, for example, require customization of the software to assure its compatibility with
other software and systems; the conversion of existing data files to allow effective operation of
the new system; the development of specialized features, functions, or interfaces; or system
maintenance and support, including periodic upgrades or enhancements to meet changing
business or regulatory requirements. Each service or component of the deal should be analyzed
separately because a misunderstanding or deficiency in any area could undermine the entire
transaction and give rise to a costly dispute.
Not only should a contract identify the different types of relationships the parties
create, it should clearly articulate their respective rights and obligations if their respective
performances fall short of expectations. In other words, the parties must address who is
responsible, and to what extent, if the goods or services delivered under the arrangement fail to
work properly, do not work at all, or are not delivered on time.
2. Managing Disputes as They Arise
Businesses frequently seek legal advice when entering into agreements and when
disputes arise. The same companies, however, often overlook the role of sound legal practice in
Page 14
monitoring and managing the implementation of their agreements. The effective management of
implementation can prevent costly disputes or improve a party’s likelihood of success should a
dispute become unavoidable. Managing problems as they arise has three main elements.
First, the parties must decide who bears responsibility for the problem. This can
be difficult. Did the problem arise because the supplier overstated the capabilities of it product?
Did the purchaser incorrectly specify its requirements? Did problems arise due to personnel
changes at either the purchaser or vendor? Or did the problem arise because external facts, such
a governmental regulations, economic conditions, or technologies, have changed.
Second, the parties should agree on a timetable for curing the defect or problem.
Third, the parties should agree on a remedy if the problem is not resolved in the
specified period. Remedies include termination of the project if the problem is critical and
modification of future fees, refunds, or other forms of consideration, such as concessions on
other projects.
All agreements regarding changes in the project or remedies for problems should
be in writing. A clear demarcation of rights and responsibilities almost certainly promotes
effective management of the project and reduces the risk of protracted disputes later on. By
contrast, failure to address significant problems can allow them to fester and grow.

3. Management of Project Disputes

(a) Escalating Negotiations

Perhaps the most common form of managing project disputes is the requirement
of escalating negotiations. Parties plan for disputes by including in their contract a requirement
that, in advance of litigation or other processes involving third parties, the parties first try to
resolve disputes through negotiation at escalating managerial levels.
Page 15
Such agreements, for example, may require each party to designate a project lead
or manager who has front line responsibility for formal communications and notices during the
course of the project. Further, the contract may specify that, prior to initiating litigation or
arbitration, a party must first bring the matter to the attention of the other party’s project manager
and then attempt to negotiate a resolution within a specified period, such as two weeks. The
contract may further require that, in the event the managers fail to negotiate a resolution, the
parties must escalate the matter to one or more higher levels of management for one or more
additional rounds of negotiation. Effective escalation requirements will, moreover, typically
require the parties to enter written amendments or change orders to formalize any negotiated
resolutions.
Escalation is perhaps most effective for managing disputes between parties that
have long standing and/or multifaceted relationships. In such cases, while front line managers
may have the best information about the specific project or dispute in question, high-level
managers may have a better understanding of the range of relationships between the parties
and/or potential tradeoffs involving matters outside the scope of the project giving rise to the
dispute. For example, offsetting cost increases on projects between the parties, with concessions
on others, may be possible.
Escalation requirements, however, are not riskless or appropriate to all business
arrangements. In some instances, such requirements may delay resolution without increasing the
likelihood of a settlement. Because of a lack of day-to-day project involvement, senior managers
may lack sufficient information or perspective about projects to fairly evaluate the dispute and,
therefore, may adopt unproductive negotiating positions.

Page 16
Further, and perhaps more significantly, an escalation provision may prevent,
rather than promote, efficient resolution of disputes because of reluctance by one or another of
the parties to involve senior managers in matters that might reflect poorly on the front line
managers themselves or others working on the project. In other words, instead of serving to
promote prompt resolution of disputes, escalation clauses sometime create incentives to sweep
problems under the rug until they grow so large that they become plainly visible.

(b) Appointment of a “Project Neutral”

In addition to bilateral processes for managing and containing disputes, it may
make sense in some instances for the parties jointly to retain a “project neutral” to mediate and
possibly even arbitrate disputes as they arise during the course of the project or business
relationship.
There are several significant benefits to the appointment of a neutral in advance of
potential business disputes. First, as noted in the prior chapter, it is likely far easier for the
parties to agree upon a neutral who has both knowledge of the subject matter of the business
relationship and a commitment to minimize dispute costs prior to the onset of conflict between
the parties. At the outset of the business relationship, both parties have incentive to minimize the
cost of potential disputes and, assuming good faith, neither is likely to have the upper hand.
Second, the appointment of a neutral in advance, and establishment of agreed
upon procedures for resolving disputes, decreases the start-up cost of seeking assistance to
resolve a dispute and, therefore, makes it more likely that the parties will act quickly to address
problems as they arise. Not only can the solution be initiated without delays for the selection of
the neutral, but it would also eliminate some of the time necessary to educate the neutral about

Page 17
the nature of the project, the terms of the contract, etc. (assuming that the neutral is provided
periodic updates on the status of the project).
(c) Controlling Dispute Costs
Despite best efforts, some disputes are unavoidable. In such cases, a party can
best manage its own potential costs of the dispute by acting quickly to prepare for litigation or
other dispute resolution processes.
A party should, for example:
• Collect relevant technical and legal documents, including all correspondence
with the other party.
• Interview project personnel.
• Prepare a chronology that describes the history of dealings between the parties.
• Fairly evaluate the strengths and weaknesses of potential claims.
• Assess potential litigation costs as part of evaluating settlement posture.
• Consider various mechanisms for resolving the dispute, including direct
negotiation, mediation, arbitration, and litigation.
In addition to evaluating the available evidence, a party to a potential dispute must
recognize that continued dealings with the other party will have legal significance. Because the
record of dealings between the parties may be murky, the manner in which the parties deal with
each other may have great significance. As a consequence, a party to a potential dispute should
identify the legal and factual basis for its position and should refrain from conduct that could be
construed as inconsistent with that position.
To mitigate damages, for example, once a decision has been made to terminate an
agreement, the termination should be done quickly to avoid the accumulation of additional
damages or business injury. The party purchasing the services should consider appointing
someone to manage all continued communication with the opposing party to avoid sending

Page 18
conflicting messages. The person in charge of such communications should work closely with
legal counsel to promote a consistent and effective legal posture for the dispute.
Finally, a party to a business dispute should coordinate its business and legal
strategies. Given the uncertainty inherent in legal proceedings involving complex matters, a
party should not rely solely on legal proceedings to resolve its strategic business concerns.

Page 19

Bibliography and Cited Works

Arkes, H., & Blumer, C. (1985). The Psychology of Sunk Cost. Organizational
Behavior and Human Decision Process, 35, 124–140.
Arkes, H., & Hutzel, L. (2000). The Role of Probability of Success Estimates in
the Sunk Cost Effect. Journal of Behavioural Decision Making, 295–306.
Armor, D. A., & Taylor, S. E. (2002). When Predictions Fail: The Dilemma of
Unrealistic Optimism. In T. Gilovich, D. W. Griffin, & D. Kahneman, Heuristics and Biases:
The Psychology of Intuitive Judgment (p. 334). Cambridge University Press.
Babcock, L., & Loewenstein, G. Explaining Bargaining Impasse: The Role of
Self-Serving Biases. Journal of Economic Perspectives, 11, 109–126.
Hastorf, A. T., & Cantril, H. (1954). They Saw a Game: A Case Study. Journal of
Abnormal and Social Psychology, 49 (1), 129–134.
Jones, E. E., & Harris, V. A. (1967). The Attribution of Attitudes. Journal of
Experimental Social Psychology, 3, 1–24.
Kahneman, D., & Tversky, A. (1982). Subjective Probability: A Judgment of
Representativeness. In D. Kahneman, P. Slovic, & A. Tversky, Judgment Bias Under
Uncertainty: Heuristics and Biases. New York: Cambridge University Press.
Knox, R., & Inkster, J. A. (1969). Postdecision Dissonance at Post Tim. Journal
of Personality and Social Psychology, 8 (4), 319–323.
Loftus, E. F., & Wagenaar, W. A. (1988). Lawyers’ Predictions of Success.
Jurimetrics, 28, 437, 450.
Page 20
Lord, C., Ross, L., & Lepper, M. R. (1979). Biased Assimilation and Attitude
Polarization: The Effects of Prior Theories on Subsequently Considered Evidence. Journal of
Personality & Social Psychology, 37 (11), 2098–2109.
Plous, S. (1991). Biases in the Assimilation of Technological Breakdowns: Do
Accidents Make Us Safer? Journal of Applied Social Psychology, 21 (13), 1058–1082.
Tversky, A., & Kahneman, D. (1974). Judgments Under Uncertainty: Heuristics
and Biases. Science, 185 (4157), 1124–1131.
Wason, P. C. (1960). On the Failure to Eliminate Hypotheses in a Conceptual