Psy/Orf 322

Human – Machine Interaction

Spring 2003

 

Helping Humans Make Better Decisions

 

 

Operations Research and formal Decision Sciences grew out of military and industrial needs to create a discipline focused on making better decisions about the challenges of

·        troop deployments,

·        re-supply strategies,

·        production processes and

·        organizational priorities, to name a few. 

 

The discipline has created computer-based tools that yield decisions about what to do, when! 

 

Some have suggested that individuals in daily life face similar, while not as daunting, decisions and could use similar tools.

 

Is this over-kill? 

 

Some have suggested that individuals in daily life face similar, while not as daunting, decisions and could use similar tools.

 

 

 

What do we do during the day:

1.      Sleep (8hrs)

2.      Work/Learning (8hrs)

3.      Eat (2 hrs)

4.      Commute (1hr +)

5.      Run errands (1hr+)

6.      Recreate/Entertainment/Learning (4 hr- + weekends)

 

What’s important:

1.   Food

2.   Shelter

3.   Clothing

4.   Travel

5.   Entertainment/Learning

 

Examples of every-day decisions for which individuals could use some help?

 

 

For example, the task of going to a particular place requires decisions by the individual of: where to go, when to leave, how to go, what route to take, turning left at the next intersection, etc.  Not surprisingly, consumer-oriented adaptations of military and industrial systems that aid personal decision making in route choice and personal finance are beginning to appear on the market. 

 

Is there a real need here, or are these simply solutions looking for a problem? 

 

Homework reported:

 

 

Greg Prisament  10:32am Monday
Article: U.S. technology can outpace decision-making, USA Today
http://web.lexis-nexis.com/universe/document?_m=7af9cc3a552067c318e6afe7291d9ccd&_docnum=2&wchp=dGLbVlb-lSlAl&_md5=5b8d90f79708b8cd72219fa488ad1f44
 

Edward Cong 3:14am  “Judgment under Uncertainty: Heuristics and Biases”

 Amos Tversky and Daniel Kahneman

 Science, New Series, Volume 185, Issue 4157, 1124-1131

 

Mati Chessin   12:22AM  http://www.mindtools.com/pages/article/newTED_04.htm

 

Jason White  "The Effects of Mood on Individuals' Use of Structured Decision Protocols" Kimberly Elseach and Pamela Barr Organization Science, Volume 10, March 1999.

 

John Andrews

Title:                 “U.S. Air Force Uses New Tools to Minimize Civilian Casualties”

Author: David A Denny, Washington File Staff, U.S. State Department

Website:           www.usinfo.state.gov/regional/nea/iraq/text2003/0318airforce.htm

 

Ryan Wells  http://www.ethics.ubc.ca/papers/invited/colero.html

 Author: Larry Colero, Crossraods Programs Inc.

Title: A Framework for Universal Principles of Ethics In use since 1997

 

Chris Asplund   Shanks, D. R.  (1995).  Is human learning rational?  Quarterly Journal of Experimental Psychology, 48A, 257-279.

 

Matt Hawrilenko   http://www.robustdecisions.com/BayesianMethodDecisions.pdf

 

Mariam Vazquez   New product decision making: How chance and size of loss influence what marketing managers see and do.

By Forlani, David; Mullins, John W.; Walker, Orville C., JR

Psychology & Marketing. 2002 Nov Vol 19(11) 957-981

 

Matthew Wolf  Determinants of Risky Decision-Making Behavior: A Test of the Mediating Role of Risk Perceptions and Propensity

Sim B Sitkin; Laurie R. Weingart

The Academy of Management Journal, Vol. 38, No. 6. (Dec., 1995), pp. 1573-1592.

Stable URL: http://links.jstor.org/sici?sici=0001-4273%28199512%2938%3A6%3C1573%3ADORDBA%3E2.0.CO%3B2-F

 

Sarah Molouki           Palma-dos-Reis, A., & Zahedi, F. (1999). Designing personalized intelligent financial decision support systems. Decision Support Systems, 26, pp. 31-47.

Tyler Mincey   Mediated reality using computer graphics hardware for computer vision
Fung, J.; Tang, F.; Mann, S.; Wearable Computers, 2002. (ISWC 2002). Proceedings. Sixth International Symposium on , 2002 Page(s): 83 –89

 

Karim Branford Machine Learning Approaches to Medical Decision Making by Konstantinos Veropoulos, Dept of Comp Sci, Univ of Bristol, Mar 2001 http://www.cs.bris.ac.uk/Tools/Reports/Abstracts/2001-veropoulos.html

 

 

ekroshus [ekroshus@Princeton.EDU]

 

Michelle Buffum  Cortes, L. (1997). Designing a Graphical User Interface. Medical Computing Todayhttp://www.medicalcomputingtoday.com/0agui.html

 

 

Benjamin P. Holskin   Gazzaniga, M., Miller, M., Wolford, G. (2000) The left hemishpere's role in hypothesis formation. Journal of Neuroscience, 20:RC 64, 1-4.

Christina Mester  Decision Support Systems Volume 32, Issue 1, Movember 2001 Pages 53-69

Building an agent-mediated electronic commerce system with decision

analysis features

Nikos Karacapilidis and Pavlos Moraitis
 

Aaron Zimmerman   Padmos & Bernstein:  Personal Web Assistant

James Ward  Gilbert, D.T & Ebert, J.E.J. (2002). Decisions and Revisions: The affective forcasting of changeable outcomes.  Journal of Personality and Social Psychology,  82(4), 503-514

John D. Alshuler  Kahneman, D. & Tversky, A. (1984). Choices, values, and frames. American Psychologist 39(4), 341-350.

Brad Bissell   "What we want to do versus what we think we should do: Empirical  Investigation of Intrapersonal Conflict" by K. O'Connor, C. De Dreu, H. Schroth, B. Barry, T. Lituchy, and M.

Bazerman Journal of Behavioral Decision Making, Volume 15

 

Mari Kobayashi  "Demystifying AI in Everyday Life" -Nick Loadholtes URL: http://artificialintelligence.ai-depot.com/Essay/Demystify.html

 

Nicole DiLello

Author: K. Papamichael, H. Chauvet, J. LaPorta, and R. Dandridge

Title:  Product Modeling for Computer-Aided Deicision-Making

Source:  Automation in Construction 8 (http://gaia.lbl.gov/BDA/documents/40110.pdf)

 

Christian A. Asmar  http://www.josseybass.com/cda/cover/0,,0471382477%7Cexcerpt,00.pdf

 

Becca Gillespie  Victor J. Stevens, Russell E. Glasgow, Deborah J. Toobert, Njeri Karanja and K. Sabina Smith.  "One-year results from a brief, computer-assisted intervention to decrease consumption of fat and increase consumption of fruits and vegetables", Preventive Medicine, Volume 36, Issue 5, May 2003, Pages 594-600

Julie Kestenman   "How Decision Makers Evaluate Alternatives and the Influence of Complexity" by Paul C. Nutt.  The article is from Manangement Science, Volume 44, Issue 8 (August 1998), pgs. 1148-1166.

Margaret Gerbasi    Ragothaman, S. & Davies, T. "Using Artificial Intelligence Techniques to Predict M.B.A. Admission Decisions" College Student Journal, March 1998, 32(1): 125-134

Garo Hovnanian    Miller et al. (2003) PDA Infectious Diseases Applications for Health Care Professionals. Clinical Infectious Diseases 36:1018-1029

Nada Siddiqui

Author: Robert Hamilton

Title: FDA Examining Computer Diagnosis http://www.fda.gov/fdac/features/795_compdiag.html

 

 

 
************************************
Greg Prisament  10:32am Monday
Psy/Orf322 Kornhauser HW
 
Article: U.S. technology can outpace decision-making, USA Today
link:
http://web.lexis-nexis.com/universe/document?_m=7af9cc3a552067c318e6afe7
291d9ccd&_docnum=29&wchp=dGLbVlb-lSlAl&_md5=5b8d90f79708b8cd72219fa488ad
1f44
 
               This article describes how war technology has helped the US
military make decisions.  Advanced satellite communications, sensors,
and weapons have made military operations very easy to control and have
allowed micromanagement by high-ranked officials.  The technology allows
for precision targeting and enemy location.
               The article states that the key problem is that, despite the
machine's aid, humans are still ultimately responsible for
decision-making.  While computers can present us with "optimal"
solutions to military problems, humans do not always go along with this
time-sensitive solution for diplomatic reasons or because of
inefficiency in the military reporting structure..
 

Edward Cong 3:14am

ORF PSY Reading for Week 10

April 13, 2003

 

 

“Judgment under Uncertainty: Heuristics and Biases”

 

Amos Tversky and Daniel Kahneman

 

Science, New Series, Volume 185, Issue 4157, 1124-1131

 

Why this article is pertinent to: “Helping humans make better everyday decisions"

 

Summary:

 

            This article tries to detail a few of the heuristics that people commonly reduce their decision making to.  The authors propose three main heuristics:  Representative, Availability, and Adjustment from anchor.  The Representative heuristic has to do with how we have a tendency to classify things based on probabilities and characteristics.  An example of this is the classification of “far” with blurry and “close” with clear.  The Availability heuristic has to do with the information that is available for recall when a decision has to be made.  An example was presented in class a few weeks ago when it was asked was the source of more deaths.  Many people chose car accident as biggest source of death in the United States because that is what they see on TV, and consequently what is available for recall.  The Adjustment from anchor heuristic states that people tend to start a base estimate than adjust from it based on previous knowledge.  This heuristic has application whenever we make numerical estimations, for example of how many countries are in the world.  This article points out some of the reasons that humans make judgment and probabilistic error in the course of everyday life. 

 

 Mati Chessin   12:22AM
http://www.mindtools.com/pages/article/newTED_04.htm
 
 
 
My article is from a company/website that is devoted to helping people
be more organized, efficient and better decision makers.  The article
discusses how Decision Tree Analysis can help people make complicated
decisions about courses of action.  It discusses how to build and use
decision trees.
 
 
lszoloma@Princeton.EDU
 
attached find the reading which I've chosen for today's class.  Also,
I'm not sure whether it's important or not, but I worked one summer on a
decision making device for the military (at Natick Army labs in MA), a
program related to packing backpack items in order to acheive optimal
load carriage by foot soldiers.  I'd like to talk about that a little as
well.  See you in class
~Lauren
 
 

 

 

Jason White

                    

"The Effects of Mood on Individuals' Use of Structured Decision Protocols" Kimberly Elseach and Pamela Barr Organization Science, Volume 10, March 1999.

 

This is a very interesting study done on the effects of temporary mood
states on an individual decion-maker's inclination to use structured
decision making processes when making complex decisions.  The particular
decision protocol used in this study was decision trees and the subjects
were business school students.  Elsbach and Barr find that individuals
in positive moods tend to skip steps in the decision making process,
execute steps out of order, and often rely more heavily on their gut
feeling.  Conversely, decision-makers in a negative mood tend to be more
vigilant in their use of decision protocol, not wanting to skip any
steps or take any chances in making a complex decision.

 

*********************************

John Andrews

ORF 322 HW 5

 

Title:                 “U.S. Air Force Uses New Tools to Minimize Civilian Casualties”

Author: David A Denny, Washington File Staff, U.S. State Department

Website:           www.usinfo.state.gov/regional/nea/iraq/text2003/0318airforce.htm

 

Summary:        

 

Civilian casualties were very high in the past partly because aerial bombing was an inaccurate science:  the circular error probability (CEP) of a WWII-era B-17’s bomb was 3,000 feet compared to 10 feet for a laser- or satellite-guided precision bomb today.  Consequently, it took 9,000 bombs in WWII to achieve the same certainty of target destruction that one bomb today can attain.

 

Nevertheless, collateral damage is still a problem – especially when exploited by Hussein et al. in “lawfare.”  The Air Force lawyers, who are concerned with international prohibitions on targeting civilians, use software to help them decide whether a target is okay to hit.  This software, called Fast Assessment Strike Tool for Collateral Damage or FAST-CD, can very quickly generate a “probable damage field” based upon type of ordnance, terrain, angle of attack, etc.  This field can then be used to estimate the impact upon the civilian population. 

 

This software is a great help to the Air Force in deciding which targets to attack with its precision-guided munitions (PGM).

********************************************************

 

Ryan Wells:

 

http://www.ethics.ubc.ca/papers/invited/colero.html

 

Author: Larry Colero, Crossraods Programs Inc.

Title: A Framework for Universal Principles of Ethics

In use since 1997

 

Abstract:

This short article contains a basic framework for ethics.  The author divides ethics into personal ethics, professional ethics, and global ethics with specific principles under each one and claims that a complete ethical code is a combination of the three.  The article is just one author's ideas; they are not the result of brilliant research or empirical testing but simply a compilation of commonly held ideas about ethics.  The framework, however, has been open to criticism from students, professors, and businessman from across the world, and only one change to the framework has ever been suggested.  Obviously, ethics are one of the many factors that influence our day to day decision making, and this article could serve as a starting point for a discussion about what ethics are, how universally they can be applied across culture or background, and how they should influence our day to day decision making.

 

Matt Hawrilenko [mhawrile@Princeton.EDU]

http://www.robustdecisions.com/BayesianMethodDecisions.pdf

 

This article describes a computer model for team decision-making.  The computer model utilizes Bayesian principles to analyze decisions with information provided by team members.  It uses mathematical formulas that weight various aspects of the decision—including subjective expected utility, marginal value of information and the probability of a decision being the best—to formulate a preference model.

 

***************************

Chris Asplund [casplund@Princeton.EDU]

 

Shanks, D. R.  (1995).  Is human learning rational?  Quarterly Journal of Experimental Psychology, 48A, 257-279.

 

This article examines the question of whether human learning is rational in the sense that it is reasoned, normative, and accurate.  Although Shanks concludes that human learning is fairly rational in all of these senses, he points to certain tasks in which assistance is certainly warranted.  For example, when attributing causality, humans tend to overemphasize positive examples and de-emphasize negative ones.  Also, they often fail to react appropriately to sample size considerations.  Both of these problems can be solved by using computational methods for deciding the causal relation between events, which can be helpful in everything from testing the efficacy of a drug to linking smoking with cancer.

**************************8

Mariam Vazquez [mvazquez@Princeton.EDU]

 

 

New product decision making: How chance and size of loss influence what marketing managers see and do.

By Forlani, David; Mullins, John W.; Walker, Orville C., JR

Psychology & Marketing. 2002 Nov Vol 19(11) 957-981

 

 

Abstract

 

  This article empirically examines, in a new-product decision context, the relationships among risk propensity, perceived risk, and risky choice decisions, when risk is operationalized as the chance of loss and the size of loss. The results indicate that perceptions of chance of loss directly influence choice among alternatives possessing different chances of loss and gain, whereas risk propensity directly influences choice among alternatives that differ in their size of loss and gain. The findings extend previous research by identifying dimension-specific effects (a) between who the decision maker is and the size of an investment's potential loss, and (b) between what the decision maker sees and the chance that an investment will experience a loss. These results not only contribute to theory, but also provide marketing managers with guidance for their risky choice decisions. The composition of a new product's risk has implications for the decisions marketing managers make,  for the placement of managers in risk-sensitive positions, and for the presentation of information to individuals with oversight responsibility for the firm's product strategy decisions.

 

 

Matthew Wolf

PSY / ORF 322

April 13, 2003

 

Determinants of Risky Decision-Making Behavior: A Test of the Mediating Role of Risk Perceptions and Propensity

Sim B Sitkin; Laurie R. Weingart

The Academy of Management Journal, Vol. 38, No. 6. (Dec., 1995), pp. 1573-1592.

Stable URL: http://links.jstor.org/sici?sici=0001-4273%28199512%2938%3A6%3C1573%3ADORDBA%3E2.0.CO%3B2-F

 

            This article focuses on the roles of factors such as outcome history, problem framing, risk propensity, and risk perception on decision making.  By asking questions about a subject’s willingness to participate in a hypothetical risky stock car race, Sitkin and Weingart attempt to confirm their hypotheses on how human’s use certain information to base their decisions.  By studying the effects of the aformentioned factors, as well as the mediating role of risk propensity and perceptions, the authors try to determine how humans base decisions on jeopardous matters.  This reading is extremely pertinent to the study of better human decision making, as the study seeks to reveal the true determinants of risky decision making behavior.

 

Sarah Molouki 

PSY/ORF 322

 

Citation:

Palma-dos-Reis, A., & Zahedi, F. (1999). Designing personalized intelligent financial decision support systems. Decision Support Systems, 26, pp. 31-47.

 

Summary:

The article discusses the issue of whether the personal characteristics of users have an impact on methods used for investment decisions, and thus whether a decision support system which takes these characteristics into account would be a beneficial tool. In order to explore this issue, the authors developed and tested a prototype system which incorporated a personalization and customization module which helped chose between different investment selection strategies based on a knowledge base of personal information gleaned from the user. It was found that users did in fact choose different methods based on preliminary characteristics such as risk aversion and gender, indicating that it may be useful to incorporate such intelligent personalization modules into investment expert systems, and perhaps decision support systems in other areas as well.

 

Tyler Mincey

Mediated reality using computer graphics hardware for computer vision
Fung, J.; Tang, F.; Mann, S.;
Wearable Computers, 2002. (ISWC 2002). Proceedings. Sixth International Symposium on , 2002
Page(s): 83 -89

 

summary:

   The term wearable computer refers to a computer device that is small and lightweight enough to be easily carried by an individual during the course of ones everyday activities. When such computers are used in conjunction with visual and auditory sensors alongside unobtrusive visual displays they can provide contextual information about ones location and activities. This paper investigates wearable computers ability to recognize significant locations and situations and to initiate context based communication channels with other users/data sources. The computer helps with everyday tasks by responding to certain environments to provide information needed to make decisions.

 

Karim Branford

PSY 322

13APR03 @ 2359

Homework 6: Machines in the Medical World

 

The article I choose was the abstract from a thesis called, Machine Learning Approaches to Medical Decision Making by Konstantinos Veropoulos, Dept of Comp Sci, Univ of Bristol, Mar 2001(http://www.cs.bris.ac.uk/Tools/Reports/Abstracts/2001-veropoulos.html).  In the abstract, the concept that machines can be used in early detection of viruses is put forth.  He claims that “the use of intelligent methods for medical decision making” has been used in automatically identifying such bacteria as tubercle bacilli from photomicrographs of spu-tum smears.  The bacteria is detected is by the use of artificial neural networks and support vector machines, and are supported by image processing machines.  Though the accurateness of such a machine is not 100% reliable, due to the small data set, the “data is still accurate enough to make a reliable decision supporting the diagnosis of the clinician.”  The further development of this technology is of great importance not only to the medical world but to all mankind (for tubercle bacilli leads to the disease tuberculosis which is the greatest killer of adults amongst all known diseases. 

 

 

ekroshus [ekroshus@Princeton.EDU]

 

With vast amounts of real-time information available, what kind of machines will help the individual make better decisions? What are the communication, computing and interface requirements? How will the supporting information be gathered and distributed. What about quality? A pragmatic example: getting from A to B, how to navigate, guide and control.

 

"Online Stock Trading: Tips everyone should read" is pertinent to the issue of human decision making by aid of computers in that it outlines some major pitfalls that may arise when humans rely on computer generated information for high-stakes activities such as stock trading.

The accuracy of the data and it's timely updating is cause for concern.

Simply because it is listed on the most progressive piece of machinery available does not mean it is continuously updated.  Further, people may falsely assume that computer transactions are executed instantaneously.

There is also an element of control loss when making decisions and completing transactions through computers, in that after accepting their hopefully accurate information, an issue of interface is that there is limited method of recourse or ability to physcially check on the transaction.  For these reasons humans may be reluctant to use computers for real-time decision making purposes, or need to at least be wary in high stakes situations.

 

 

Homework #6 for Michelle Buffum

 

Cortes, L. (1997). Designing a Graphical User Interface. Medical Computing Today

http://www.medicalcomputingtoday.com/0agui.html

 

One of the most important keys to a machine that is a decision-making aid is that the user can understand how to use it. Leslie Cortes’ article describes techniques for making programs that are functional and have an interface that works well with its users. The interface is important because it determines the navigation of a site. Grandma may need health advice, but if she can’t figure out how to find her problem on a web site then it’s good intentions and wealth of knowledge is worthless to her. The information a site has is only useful if the interface is designed in such a way that people can easily and happily figure out how to access it.

 

 

Benjamin P. Holskin [bholskin@Princeton.EDU]

 

Gazzaniga, M., Miller, M., Wolford, G. (2000) The left hemishpere's role in hypothesis formation. Journal of Neuroscience, 20:RC 64, 1-4.

Humans have the tendency to jump to conclusions about a variety of circumstances, especially situations involving correlated events.  This study suggests that the left hemishpere of the brain plays a significant role in our tendency to equate correlation with causation.  By testing split-brain participants on a guessing task, one in which humans often report finding a pattern when there is none, the researchers were able to show that while the right hemishpere may be more "reasonable" (i.e. not jump to conclusions), the left often automatically searches for a causal relationship.  Since the left also controls language and communication, these predicted causal relationships manifest themselves more often than the more conservative conclusions drawn by the right.  From an evolutionary/survival perspective, there are certain scenarios where it makes sense for humans to jump to conclusions: you're in the jungle, you hear a roar, you automatically assume it's a tiger and you run.  In other, more subtle situations it helps to take into account both the causal conclusion as well as the more conservative assumptions "proposed" by the right hemisphere.

 

Christina Mester

 

Decision Support Systems

Volume 32, Issue 1, Movember 2001

Pages 53-69

 

Building an agent-mediated electronic commerce system with decision

analysis features

Nikos Karacapilidis and Pavlos Moraitis

 

The authors of this paper designed and implemented a computer based

system that helps facilitate e-commerce. Their system provides _agents_

for both the seller and the buyer, which help them sort through offers

and make purchasing decisions. Somebody who wants to buy something can

_hire_ an agent program and give it specifications and preferences for

the product he wants to purchase. Similarly, the seller has an agent,

which then communicates with the buyer_s agent. After collecting a

number of offers, the buyer_s agent then makes a recommendation for what

product should be purchased. Both buyers and sellers can update their

preferences at all times. Graph algorithms are used for some of the

decision making by the agents. The interface with the clients is

internet based.

Aaron Zimmerman [azimmerm@Princeton.EDU]

 

Padmos & Bernstein:  Personal Web Assistant

 

This article is a very interesting article written 6 years ago here at Princeton University. I like the fact that is was written a while back because we can see the thought process of academics determining the viability of technology while knowing how that technology is performing today! The papers focus is on various aspects of Personal Travel Assistants

(PTA¹s) such as: continuous connection range, proposed protocols and potential data sources. They explain the reasoning behind their main conclusion (that the WWW provides a good potential data source) nicely; and it is fitting that at this point in time a lot of wireless devices do in fact (I believe) get their information from the WWW. They say the functionality of these devices to and their ability help humans make better decisions is just about limitless and I think this will spur good in class discussion.

 

James Ward [jward@Princeton.EDU]

Gilbert, D.T & Ebert, J.E.J. (2002). Decisions and Revisions: The affective forcasting of changeable outcomes.  Journal of Personality and Social Psychology,
82(4), 503-514.

This article explains how humans can make decisions that they believe will make them happier but actually make them less happy in the long run.  It found that, when people are given the option to make a "changeable" decision, they believed they would be happier with this decision than if their decision was not changeable.  However, when they were actually tested, they were happier with the decisions which they were not given the option to change in the future.  This result shows that we as humans believe that a decision with more option is better when this might not be the case.

John D. Alshuler [alshuler@Princeton.EDU]

Kahneman, D. & Tversky, A. (1984). Choices, values, and frames. American Psychologist 39(4), 341-350.

 

This paper is the basis of prospect theory, which is descriptive theory of how we make decisions, and the framing effects, which means that we make different decisions depending on how the question is phrased.  For example we are loss-averse, so a loss of $100 is more painful than a gain fo $100.  Another example is of how the answers are framed.  Therefore, if we had to pick a cure for a disease, we would choose saving lives, rather than gambling, but we would rather gamble than take a sure loss.  This shows that we are risk-seeking for losses and risk-averse for gains, when in actuality, when in actuality, the cure is the same thing, we just answer differently.  The framing effect and prospect theory has huge implications about how we can make better everyday decisions.

 

 

Brad Bissell [bbissell@Princeton.EDU]

 

"What we want to do versus what we think we should do: Empirical

Investigation of Intrapersonal Conflict"

by K. O'Connor, C. De Dreu, H. Schroth, B. Barry, T. Lituchy, and M.

Bazerman

Journal of Behavioral Decision Making, Volume 15

 

            People are able to distinguish between what one should do and what one

wants to do during the decision-making process.  This article explains

that humans often make decisions even though they know the decision is

not one that they should make.  Therefore we do not always make the

most moral or fair decision because we want to maximize our own

benefits and pleasures.  Human decision-making can be biased and has

many subjective qualities that a computer would not have.  Computers,

being unbiased, would make objective decisions that maximize joint

rewards.  The use of computers for making decisions in certain

situations is definitely supported by this logic.

 

 

 

Mari Kobayashi

 

"Demystifying AI in Everyday Life" -Nick Loadholtes

URL: http://artificialintelligence.ai-depot.com/Essay/Demystify.html

 

This essay attempts to show that AI can provide computers with a set of principles for decisions-making in a fashion similar to humans. Such AI systems gather information, filter it discarding irrelevant details, and produce an answer suited to solving the problem at hand.

 

Here is an excerpt from the abstract:

"The goal of the field of Artificial Intelligence (AI) is to produce systems (be they robots, computer programs, user interfaces, etc.) that utilize behaviors and decision making skills that are similar to ones that humans (and other living creatures) use. In essence, to create or mimic intelligent behavior in non-sentient entities. The wealth of knowledge that has been gathered in researching this topic reveals much about logic, perception, and common sense. This essay intends to show that the principals behind AI are not out of reach to the common persons; in fact, these principals can be utilized in many common situations to highlight the basics of human thinking."

 

 

Nicole DiLello

PSY322

HW #6

Due: April 14, 2003

 

Author: K. Papamichael, H. Chauvet, J. LaPorta, and R. Dandridge

Title:  Product Modeling for Computer-Aided Deicision-Making

Source:  Automation in Construction 8 (http://gaia.lbl.gov/BDA/documents/40110.pdf)

 

This article describes a new type of software that was developed to aid architects and engineers.  Previously, when constructing buildings, the architects relied on two different types of software programs.  They used a CAD program to design the building geometrically and a simulation tool to design the physical aspects, such as energy and comfort levels.  The program that was developed is a way to combine both of these programs to quicken decision making process.  The impetus for this was the energy crisis in the 1970s.  The designers set out to develop “cost-effective and environmentally friendly strategies and technologies to improve the energy efficiency of buildings without compromising comfort.”  Thus, this new technology could help engineers and architects make better decisions based on simulations instead of actual testing.  This helps them save time, money, and energy.

 

Christian A. Asmar

 

http://www.josseybass.com/cda/cover/0,,0471382477%7Cexcerpt,00.pdf

 

This reading is an excerpted chapter from Wharton on Making Decisions by Stephen J. Hoch (Editor), and Howard C. Kunreuther (Editor).  The reading discusses about the complex nature of the decisions people (specifically company managers) make and gives a list of pitfalls and some ways in which you can avoid them.  It also goes into the theory of what we know about decision making and spends some time discussing the different models of choice (normative or descriptive) and giving examples of each.  The reading focuses primarily on how humans can make better everyday decisions as managers in the workplace.

Becca Gillespie

I chose a reading about people using computers to help them make decisions about what to eat.  This is an example of an everyday decisions where computers can be helpful because a computer can store all the nutritional information easily and calculate excesses/deficiencies in the diet. Victor J. Stevens, Russell E. Glasgow, Deborah J. Toobert, Njeri Karanja and K. Sabina Smith.  "One-year results from a brief, computer-assisted intervention to decrease consumption of fat and increase consumption of fruits and vegetables", Preventive Medicine, Volume 36, Issue 5, May 2003, Pages 594-600

Julie Kestenman [jkestenm@Princeton.EDU]

    Hi. The article I read for the week was "How Decision Makers Evaluate Alternatives and the Influence of Complexity" by Paul C. Nutt.  The article is from Manangement Science, Volume 44, Issue 8 (August 1998), pgs. 1148-1166.
    This article investigates the types of evaluation tactics that managers use to make decisions for their companies. I think that the article has some important analogies to personal decision making, including the use of judgement tactics.  The result that I found most intersting from the study was the negative affect that "expert views" had on the success of a business decision.  Another finding was that decision makers used analysis more often than judgement as the number of alternatives for a decision increased. These results may provide some insight into how people formulate everyday decisions when faced with many choices and how recommendations from authority figures and friends influence personal decision-making.

 

Garo Hovnanian

Miller et al. (2003) PDA Infectious Diseases Applications for Health Care Professionals. Clinical Infectious Diseases 36:1018-1029.

There is a growing movement to use PDA's as an integral part of the medical diagnosis process. The doctor, using the information stored in an application on his PDA, can input a patient's current symptoms and medical history into the software, and get a recommendation for diagnosis and treatment, including drug names and dosages. In addition, the patient's history can be stored for easy reference in future evaluations. Adding this machine-controlled element to the diagnosis and treatment process is incredibly useful in helping to prevent a lack of foresight in diagnosing a less-common disease, the prescription of an incorrect drug that will unintentionally interact with the patient's other symptoms, etc.; in general, by calling on a medical database from their PDA's, doctors and other health professionals (like EMT's, chemical warfare specialists, and hospital residents who don't necessarily have as much experience) can minimize human decision error at crucial times like diagnosis. These programs, however, differ from designer to designer, using different priorities to assign the recommended diagnosis and treatment. So, doctors have to review the available software and determine which is best suited for them. Also, there are issues of privacy and security, limited storage capacity of the PDA, the danger of breaking the PDA and losing information, and limited Internet access, which is helpful (if wireless) to update the hospital's main database with all the information the doctors are gathering.

 

 

Nada Siddiqui

ORF/PSY 322-Pset 6

 

Author: Robert Hamilton

Title: FDA Examining Computer Diagnosis

Source: FDA website

Webpage:http://www.fda.gov/fdac/features/795_compdiag.html

 

Computer diagnosis, specifically medical diagnostic software, has become one of the most important tools in medicine. Though they do not replace the doctor’s role in diagnosing disease and probably cannot replace human characteristics such as judgement based on years of experience, intuition or the ability to make leaps of logic or to consider each patient’s personal situation, computers provide the memory to store entire medical libraries. After being given symptoms, medical histories and test results, they use these vast stores of information to suggest diagnoses of rare illnesses that a doctor may never have seen before or may not otherwise have considered or provide several possible diagnoses that a doctor can then examine further. Computers have also objectified computations needed in diagnosis eg. Psychological assessments and are being used to identify abnormal pap specimens or pre-cancerous cells that the medical technician may have missed. Recent developments have improved the accuracy and ease of use of such diagnostic software with hardware requirements reducing to a desktop and modem/cable and memory capacity increasing.

Margaret Gerbasi [mgerbasi@Princeton.EDU]

Ragothaman, S. & Davies, T.

"Using Artificial Intelligence Techniques to Predict M.B.A. Admission Decisions"

College Student Journal, March 1998, 32(1): 125-134

 

This article explains how machine learning techniques can be trained and used to make decisions about admission to business school.  The program used, IXL, was compared with other statistical models and was found to be as accurate at predicting acceptance and rejection from business schools (about 90% correct).  IXL can help admissions deans make better decisions by classifying applicants and being flexible enough to weigh various factors with uncertainty so that the classifications that are made can be easily reviewed.  IXL can also help by removing some of the bias involved in admissions and by being more accommodating to a change in the rules governing admission than humans.

Mati Chessin [mchessin@Princeton.EDU]

 

y article is from a company/website that is devoted to helping people be more organized, efficient and better decision makers.  The article discusses how Decision Tree Analysis can help people make complicated decisions about courses of action.  It discusses how to build and use decision trees.

 

The link:   http://www.mindtools.com/pages/article/newTED_04.htm