Scribd
Upload
17 views

Judgmental Methods

Judgmental forecasting methods rely on human judgment, typically from experts. While commonly used, experts often produce poor forecasts due to overconfidence in their abilities. Data-driven methods have been shown to produce more accurate forecasts than experts. It is important for experts not to be overconfident and to consider outside perspectives when making judgments.

Uploaded by

Vikash Movva
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
17 views

Judgmental Methods

Judgmental forecasting methods rely on human judgment, typically from experts. While commonly used, experts often produce poor forecasts due to overconfidence in their abilities. Data-driven methods have been shown to produce more accurate forecasts than experts. It is important for experts not to be overconfident and to consider outside perspectives when making judgments.

Uploaded by

Vikash Movva
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 14

Forecasting Demand – Judgmental Methods

Anton J. Kleywegt, Ph.D.


School of Industrial and Systems Engineering
Review of Previous Lessons

• Careful thought is required when modeling


demand
• Customer behavior often is not rational
• Customer behavior exhibits predictable
patterns even when it is not rational
Learning Outcomes
• Become familiar with some characteristics of the most
widely used method of forecasting: judgmental
methods
Forecasting Methods
• Classification of forecasting methods according to Armstrong, J.S. Long-range
Forecasting, Second Edition, 1985
Judgmental Forecasting Methods

• As the name indicates, judgmental forecasting methods are driven mostly by human
judgment, typically the judgment of experts
• By far the most widely used forecasting approach in practice
• Examples
• Sales managers forecasting future sales or demand
• Medical doctors diagnosing causes of medical problems
• Medical doctors predicting prognoses for patients
• Pundits forecasting outcomes of elections
• Brokers recommending investments
• College admissions staff forecasting academic success
Judgmental Forecasting Methods

• “Expert” opinion often produces poor forecasts


• Quality of forecasts tends to increase and then
decrease as the level of expertise increases
• See the following articles:
• Armstrong, J.S., “How Expert Are the Experts?”,
Inc, pp.15-16, 1981
• Armstrong, J.S., “The Seer-Sucker Theory: The
Value of Experts in Forecasting”, Technology
Review, pp.16-24, 1980
Judgmental Forecasting Methods

• Proposed explanations for poor forecasting


performance of experts
• Experts think they know better, and do not learn
as well from data as more humble forecasters
• Experts are overconfident, overestimate their
knowledge and forecasting skills, and
underestimate their forecasting errors
Judgmental Forecasting Methods
• Overconfidence are positively correlated with various
factors:
• The higher the publically perceived level of
expertise, the greater the tendency to be
overconfident
• Most managers and companies perform
poorly after their praises have been sung
in the popular media
• Past success greatly increases overconfidence
• If a person is successful, it is usually
ascribed to some inherent talent, and if
a person is unsuccessful, it is usually
ascribed to external influences or bad
luck
Judgmental Forecasting Methods
• Overconfidence are positively correlated with various
factors (continued):
• Expectations of skill is positively correlated with
overconfidence
• The higher a manager’s level in the
organization, the more the tendency to
be overconfident (note that either may
be the cause of the other)
• Gender: men have a greater tendency to be
overconfident than women
• Do you think that you are a better than
average driver?
Judgmental Forecasting Methods

• Organizations continue to use experts for their


forecasting even if they know that experts do not
produce better quality forecasts
• Why?
• The manager knows that the forecast is always
wrong
Judgmental Forecasting Methods

• Judgmental forecasting methods perform poorly


compared with data-driven forecasting methods
• Even very simple, crude regression models produce
better forecasts than experts
• See the following articles:
• Dana, J. and Dawes, R.M., "The Superiority of
Simple Alternatives to Regression for Social
Science Predictions", Journal of Educational and
Behavioral Statistics, pp.317-331, 2004
• Dawes, R.M., "The Robust Beauty of Improper
Linear Models in Decision Making", American
Psychologist, pp.571-582, 1979
Judgmental Forecasting Methods
• Judgmental forecasting methods perform poorly compared with data-driven
forecasting methods
• See the following articles (continued):
• Kahneman, D. and Tversky, A., "On the Psychology of Prediction“, Psychological
Review, pp.237-251, 1973
• Lovallo, D. and Kahneman, D., "Delusions of Success", Harvard Business Review,
pp.56-63, 2003
• Overconfidence inhibits use of the “outside view”
• Example of “outside view”:
• Most entrepreneurs who start a new business forecast that it will be
successful with high probability, but fewer than 20% of startups
survive more than a few years (forbes.com)
Judgmental Forecasting Methods

• Judgmental forecasting methods perform poorly


compared with data-driven forecasting methods
• See the following articles (continued):
• Meehl, P.E., "Causes and Effects of My Disturbing
Little Book", Journal of Personality Assessment,
pp.370-375, 1986
• Sarbin, T.R., "A Contribution to the Study of
Actuarial and Individual Methods of Prediction",
American Journal of Sociology, pp.593-602, 1943
• Sarbin, T.R., "Prediction and Clinical Inference:
Forty Years Later", Journal of Personality
Assessment, pp.362-369, 1986
Summary
• Judgmental forecasting methods are driven mostly by
the opinions of experts
• By far the most widely used forecasting approach in
practice
• Judgmental forecasting methods almost always
perform poorly compared with data-driven forecasting
methods
• It is important not to be overconfident, and to take
into account the outside view

You might also like