Decision-Making Master Ralph Keeney Answers Your Questions


We recently solicited your questions for Ralph Keeney, a decision analyst at Duke University’s Fuqua School of Business.

In his answers, you’ll find that Keeney discusses how to avoid making the wrong decisions, how to figure out what you really want, and why neither psychologists nor economists have definitively figured out how to make good decisions.

Thanks to all of you for the good questions and especially to Keeney for his participation.


There’s a gap between making a decision and actually doing something about it. People can easily avoid following through on any decision that they make, and rationalize their lack of effort afterward. Given that, why is “decision making” relevant to the way people live their lives?


Decision making is relevant because it offers you the only way to purposefully have an effect on your life or on anything else. The rest of life just happens due to circumstances beyond your control, including the decisions of others. Of course, you must follow through on your decisions to have any effect, and people often do not follow through. When an individual recognizes that not following through is having a negative impact on his or her quality of life, it might be useful to create alternatives that make it easier to follow through. For example, you can announce your decision to others who will help you follow through, or you can keep your home free of junk food once you decide to cut down on snacking.


Is there a way to push other people to make decisions quickly?

I’m often troubled by my friends who can’t make decisions on simple things like the date and time for a meet-up or dinner. They often ask me to block off three to four possible days for them, making it impossible to optimize my schedule.


You can sometimes influence the decisions that others make, but you cannot guarantee that they will make them to your satisfaction. The way that you can influence them is through the decisions you make. Some of these may involve requests or logical explanations of why it might be better for your friends to make decisions more quickly. Since you totally control your decisions, other choices may be more practical for you. If you are asked to block off three or four days, you could simply say no and offer to block off either one or two days. You could separately or additionally set a time limit for the block. Alternatively, you could tell your friends that you will inform them when the block ends, and that this will happen when other plans are arranged for the same time period.


Thanks for taking some time out to help others (myself included) who are unable to make their minds up about something or other. I would consider myself to be a pretty rational person, however I am really struggling with this one.

I am originally from Ireland. I moved to England to go to university and now work in London full time. I have had a long-distance girlfriend (since I moved to university) who is currently studying medicine in Ireland. She has another two years until she can move to London to join me. That would be seven years apart in total. I have an opportunity to move back to Ireland now, where I think I would have more friends, family, and obviously my beloved.

However if I wait it out two years she will move over here to be with me, where there are much greater opportunities in mine and her line of work and substantially higher salaries on offer. (I am already earning around 50 percent more in London than I would if I moved back to Ireland.)

I have tried to use my professional skills on this and considered this as an NPV calculation (!), with short term investment for considerable payoff in the future. However I am unable to remove the emotional attachment I have to “home.”

Can you please provide some sort of third-party analysis which may help me make the best decision for all involved.


To legitimately offer advice on a significant decision like this, a meeting and a dialogue would naturally be more appropriate and useful. In such, I could help you clearly articulate all of your objectives and values, personal and professional, that are relevant, assess your preferences for achieving each of these, generate a full range of alternatives, and then evaluate these alternatives. Nevertheless, with the information in your question, I will make a few observations.

You state that your girlfriend could get a good job in London in two years. I assume if you moved to Ireland now, you could also move back to London in two years and get a good job. Hence the decision mainly involves the next two years.

How you spend any two years of life is important and worthy of careful thought. For this decision, your objectives should basically be related to the quality of your life. Components of this quality would likely include your personal life (fun, excitement, family), your professional life (interest, challenge, contribution, and salary), the quality of life of your girlfriend, and perhaps contribution to Ireland. Regarding alternatives, it may be possible to create some alternatives where you can work for your current employer partly in Ireland. Possibilities include a three-month stint or an arrangement that allows you every third Monday out-of-the-office, so you could have three-day weekend visits to Ireland at least once a month. In any case, 20 visits per year for weekends would likely not use up all of the 50 percent greater pay of your London employment.

Compare any alternative that you can create with your “stay in London” alternative and the “move to Ireland” alternative. For each of these alternatives, include the choice in two years of whether to live in Ireland or London. Finally, the NPV (net present value) calculation that you mentioned is only relevant to your financial objective. With two professionals (you and your girlfriend) it would seem likely that any difference in NPV for a two-year period would be rather inconsequential relative to all the personal concerns. Hence, in this decision, you may wish to weigh your personal concerns more than your financial concerns. I hope these thoughts are helpful.


What do you think was the decision making process of David Letterman or Tiger Woods regarding the way they conducted themselves with women? In other words, what were they thinking, and why did they think they could get away with it?


Any comments that I might make about what individuals were thinking when they made personal decisions would be pure speculation. However, I can comment on circumstances that frequently lead to poor decisions. By poor decisions, I mean those decisions for which you subsequently ask yourself, “how could I possibly have made that choice”. The basic reason is often that you did not think about your decision or that you did not use relevant information that was available to you. This often leads to one or more of four fundamental decision making mistakes. These are

1. Not recognizing, or not properly taking into account, one or more of your relevant objectives that is affected by your choice,

2. Not identifying, and therefore not considering, an alternative that subsequently came to your mind and was then easily recognized as a much better alternative than the one you had chosen,

3. Not recognizing, or not properly taking into account, possible future events that might occur, and subsequently did occur, that led to unanticipated undesirable consequences,

4. Only considering short-term consequences of the alternatives and neglecting any possible long-term consequences.

You can lessen the likelihood of all of these mistakes if, before choosing an alternative and ‘moving forward’, you give serious thought to identify your objectives, generate a full range of alternatives, consider possible long-term consequences of your alternatives, and imagine possible events that could significantly influence the eventual consequences of your choice. In short, think before you act, which for your decision making means, think hard before choosing an alternative when facing an important decision.


What questions should you always ask yourself before you make a decision – is there a checklist or a rule of thumb?


A few years ago, John Hammond and Howard Raiffa, two colleagues from Harvard, and I wrote a book called Smart Choices. In it, we tried to synthesize the key ideas to make good decisions that have been developed over time by many in the decision sciences field. Our structure included eight steps:

1. Identify your real decision problem.

2. Specify your objectives.

3. Create a full range of alternatives.

4. Understand the consequences of the alternatives.

5. Make explicit the inherent value tradeoffs.

6. Clarify the relevant uncertainties.

7. Account for your risk tolerance.

8. Consider implications for interrelated decisions.

You can read a free description of each of these eight steps by finding the book Smart Choices on Amazon, then moving your cursor over the picture of the book on the left, and clicking “First Pages” when it appears.


How can I avoid paralysis by analysis?


Paralysis by analysis refers to repeatedly delaying a decision to allow for more analysis to make a better choice. More analysis costs money and the delays use time and likely cause psychological wear and tear (e.g. stress). Depending on the decision context, the delay could allow competitors to beat you to an opportunity or a desirable alternative that you are considering could disappear (e.g. a house that you are considering buying is sold to someone else). Recognize all the negative consequences of waiting to decide and compare these to the likely improvement of delaying the decision. This comparison should reduce paralysis by analysis.


At what point do you know that you have enough information to make a rational decision?


I define a rational decision as one that you make consistent with the information you have and your values. Hence, you can make a rational decision at any time you must. A related decision is whether to gather more information so that you can possibly make a better decision. If you are considering how many units of a new product to manufacture, a market survey of potential customers may provide useful insight about how many units may be sold. To conduct the survey will cost money and require time. There are well-defined procedures in the decision analysis literature to explicitly address whether the additional information is worth that time and cost. Such an analysis indicates whether you have enough information, meaning the appropriate amount, to make the decision.


Is there really wisdom in the crowds? Do crowds make good decisions?


Making a decision involves consciously choosing among alternatives, though not necessarily with the aid of any analysis, and then taking action to implement that alternative. This is different than just taking action without considering alternatives, and perhaps without thinking. So I don’t think crowds make decisions. Individuals, and organized groups within crowds, can make decisions. Certainly, an individual or a group can attempt to steer a crowd to act in a way that the individual or group desires, and that action may be in the individual’s or group’s interest only and not in the interest of most of the people in the crowd. On the other hand, it is often the case that a group will arrive at a better decision together, as individuals may add important factors that other individuals have overlooked.


Presumably there is both an art and a science to your method of decision making. On the science side, when you model different outcomes to determine how comparatively good or bad they all are, you have to attach some kind of value to every outcome.

How do you do that? How would you value something like a marriage, or something similarly personal and subjective? Even an uncannily analytical client wouldn’t be able to attach a numerical index to something like that, right?


All decisions have consequences, which are also referred to as outcomes. Our basic intent in making decisions is to cause consequences that are improvements over the current situation and to avoid consequences that result in a degraded situation. Almost everybody will face the issue of complex consequences in some of their important decisions. The reason that one cares about consequences is based on his or her values, which are of course subjective and personal. Parenthetically, life would be incredibly boring if there were objective guidelines for personal values that would essentially eliminate individual preferences for consequences that affect them.

When facing decisions with possible consequences that include many difficult-to-measure implications, I feel it is often useful to aid your intuitive thinking with systematic analytical thinking. Furthermore, in these cases it is often important to include personal, social, health, and/or environmental consequences in those analyses. An entire area of decision analysis, called multiattribute utility analysis, has been developed for this purpose. When such an analysis is done well, it can provide important insights and understanding. And naturally, when it is done poorly, it adds nothing and is a waste of time.

A couple of examples may be useful. I have worked on the evaluation of potential sites for the long-term storage of nuclear waste from power plants. Aside from economic costs, the analysis explicitly addressed possible health and safety impacts, environmental consequences, social consequences to people in communities near any selected site, and visual and archaeological consequences. For a very different type of significant decision — namely, when should a professional woman have her first child? — Dinah Vernik, then a doctoral student at Duke and now a professor at Rice University, and I developed an analytical tool to integrate the personal, professional, and family aspirations of a woman desiring a professional career. We applied this model with several young women facing this decision who felt that it was important to understand all aspects of their choice about when to have a child. A short video description of this study is found here.


I find I have the most difficulty figuring out what it is that I want. This turns my decision-making process into an exercise in guessing at my preferences. Do you have any advice on how one can more easily discover his or her preferences?


Knowing what you want is absolutely critical to making decisions. So if you don’t know what you want in a particular decision situation, your initial effort should be to find out. The good news is that you have access to the source for all of your preferences; it is your mind. The subject of value-focused thinking is concerned with helping you decide what you want (i.e. your preferences) in any decision situation. It includes processes such as creating an initial wish list of what you care about and expanding this by considering good and bad aspects of real or hypothetical alternatives, asking others what should matter, and listing the goals or constraints that you care to include. Next you push your thinking by asking for each of the items on your list, “Why do I care about this?” The answers should expand your list. Then you convert each item to a common form called an objective, using a verb and an object (e.g. have fun, make my spouse happy, increase my salary, lose weight). For your decision, you then consider which of the objectives are most relevant and compare the alternatives on those objectives. The decision will involve making value tradeoffs (discussed in the next question) to select the alternative that best balances the achievement of your competing objectives.


Many important personal and business decisions seem to come down to the proverbial balancing of apples and oranges. It doesn’t seem right to just ask whether apples or oranges are more important and then choose the alternative that gives you the most of the chosen fruit. Often we want both apples and oranges. So is there anything that we can do to think about the decision in some way that makes sense out of this situation?


One of the most significant errors in decisions made by business and government executives, as well as individuals, is the incorrect balancing of achievement on different objectives. To illustrate this issue, I have asked numerous individuals in government, the news media, businesses, and universities the following question: “In cleaning up hazardous waste sites, what is more important: the economic cost of cleanup or the health consequences?” Everyone answers, and almost always the response is health consequences. Then I asked them this question: “In cleaning up hazardous waste sites, what is more important: reducing the health consequences from 20 to 10 people who get seriously ill for one week and then recover or reducing in the economic cost of cleanup from $2 billion to $1 billion?” Now, almost all of these same individuals respond that the economic cost is more important. I asked them how is it that the health consequences are initially more important and yet the economic cost is more important in the second situation. The response is invariably that it is not worth $1 billion to avoid the 10 serious illnesses. The general point is that asking whether economic cost or health consequences are more important is a very poorly posed question. Any informed response depends on knowing how many and the seriousness of the health consequences and knowing what the economic costs are. Otherwise, we do not know what we are expressing value judgments about.

For reference regarding this particular value tradeoff, our country could prevent more than 100 fatalities for each additional $1 billion invested in highway safety, so it would not seem socially or individually desirable to spend $1 billion to avoid 10 one-week illnesses from any hazardous waste exposure.


Something has been bothering me for a time now. Why do both psychology and economics researchers both claim to have cracked how to make good decisions, yet offer very different solutions? How does the economics/statistics aspect of decision-making deal with the role of irrational emotions in decision making?


You raise some very interesting questions. However, I don’t think that either psychological or economic researchers have definitively found out how to make good decisions, nor do I expect that this will be occurring anytime soon.

It is worthwhile to briefly explain the three main approaches for studying decision making.

Normative researchers, which include many economists, investigate how individuals should make decisions that are consistent with certain specified logical criteria. Regarding investment in stocks, the criterion stated may be to maximize the net present value of the investment. This neglects the concern of many investors to include other criteria like “promote green companies” or “avoid tobacco and gaming stocks.” I feel that if an investor has these values, it is completely rational to invest consistent with these.

Descriptive researchers, which include most psychologists and behavioral economists, are concerned with how individuals actually make decisions. They have identified numerous “decision traps” that lead individuals to make decisions that are not consistent with that individual’s stated values or information. Decision traps refer to ways that our minds think through certain complex aspects of decisions that cause us to neglect important considerations. It is very hard for any of us to think through all of the aspects of a complex decision using common sense without any systematic analytical assistance.

Prescriptive researchers and practitioners, the group in which I fit, are concerned with helping individuals and organizations make good decisions. Because it is almost impossible to consider informally all important aspects in a complex decision well, prescriptive practitioners use models and systematic procedures to provide insight to complement intuition and help make more informed choices. I often feel that the key contribution of such efforts is to avoid poor alternatives more than it is to necessarily choose the best alternative. If the decision maker were always selecting among the set of best alternatives available, he or she would likely be doing very well, whereas one very poor choice could have disastrous consequences.


For significant decisions made by many of our best-known corporations, computers will grind away on data to provide answers. Often we find out somewhat later that the answers weren’t really very good. With hindsight, the problem is recognized as having been caused by “garbage in, garbage out.” Yet, the “garbage in” was well camouflaged and appeared to be highly reliable, accurate information. Can this situation be addressed logically, and if so, how?


Decision makers often feel that collecting all the data relevant to a decision is a necessary initial step in making an informed decision. Data collection, which may be expensive and require a lot of time, sometimes begins before the decision being faced is well understood and the alternatives are identified. Hence, the collected data are often not particularly relevant to the decision at hand. That provides the “garbage in” from which the “garbage out” follows. Data are useful for a decision if they might change the alternative that should be selected and if the expected implication of that change is significant enough to justify the expense and time required to collect the data.

Suppose a small firm produced rooftop solar heating panels and needed to decide whether to build a major expansion in capacity. They could rush off and collect data on how many rooftops in the business area are appropriate for solar products and also conduct extensive consumer research about possible purchase behavior. It may be that the much more important information for the decision is whether the state legislature would pass a bill subsidizing purchases of such solar heating panels and what the price of heating oil might be in the next few winters. Carefully assessed judgments about these two uncertainties would be much cheaper and more relevant to the capacity expansion decision than the extensive data collection on rooftops and possible purchase behavior, which could be a waste of time and money.


My (very limited) understanding of game theory would tell me that when presented with a choice I should choose the “mini-max solution” — that is, choose the course of action for which the worst outcome is least bad.

First, have I correctly understood what game theory has to say about the mini-max solution? Second, does your research tend to support the wisdom of pursuing the mini-max solution, or does it turn out (as I suspect) that to do so ends up being overly conservative in many cases?


The mini-max solution is a fundamental result of game theory, which originated with analyses of two-person games. Decisions consistent with this solution require choosing the alternative that minimizes the maximum possible loss. Your question seems to concern your decisions independent of the game context. For decisions outside of a game context, I believe the preferences of most individuals are not consistent with the mini-max solution. Rather, it is better to specify your relative preference (i.e. utility) for all possible losses and gains and combine these with their likelihoods of occurrence when evaluating potential alternatives. To illustrate, consider that you could invest $10,000 in one of two alternatives: alternative A will result in either a gain of $10,000 with a one-half chance, or otherwise a loss of $2,000. Alternative B will result in either a gain of $1,500 with a one half-chance, or otherwise a loss of $1,000. Alternative B is the choice consistent with a mini-max solution, as the possible $2,000 loss in alternative A is worse that the $1,000 loss in alternative B. However, many individuals would likely prefer alternative A to alternative B, as the increase in gain from $1,500 to $10,000 is more significant than the increase in loss from $1,000 to $2,000.


I read the first third or so of this Q&A and it seems to me that Mr. Keeney presumes a degree of rationality to human behavior that just isn't so.


I also read the first third of this Q&A but couldn't read anymore - Mr. Keeney needs to make his work a bit more accessible and describe things in a more exciting way.

Ross Macaulay

Thanks Ralph for the great answers (particularly to the question that I posted!)


The two posters above make a deep flawes analysis of Dr. Keeney life long work.

The motivation of Dr. Keeney's work and other Decision Scientists like him stems from the rationality limits of the human decision maker. Prescriptive decision analysis is all about helping rational bounded humans to make better decisions by allowing for a systematic analysis towards eliminating the natural decision bias people face.

I would strongly recommend for the two posters to read some of the literature on prescriptive decision analysis before commenting.


No one taught me to bring a rational framework to the act of making life-altering decisions. The outcome of thoughtful decision-making would certainly be more beneficial than the random process I used for many years. This information was very accessible, even though I know nothing of the science or the field. The concept of improving people's lives by helping them understand how to make better decisions is, in itself, and exciting concept. It doesn't require a big government program to affect change. Instead we only have to reframe our thinking about how and why we do things and whether or not those actions will help us achieve the outcomes we desire.

Howard Raiffa

I work closely with Ralph. He does as he says he should do and consequently he does things that are more dramatic in life style than the rest of us,including me. This accounts for him living in San Francisco and working thousands of miles away, Or giving up a tenured appointment at MIT because it would interfere with skiing and surfing. He thought it through using multi-attribute decision analysis and then listened to what the analysis was telling him.
He believes what he preaches. I think he's great.

Kirk E

Wel, I think I have it figured out. I will just call Ralph to help me with my major decisions.

Kirk E

Paula Ecklund

I work where Ralph works and occasionally have the pleasure of chatting with him. Ralph seems to be a very happy person. I suspect that's at least in part because he's made (and continues to make) a careful study of what he really wants and then makes the decisions that get him those things. Also - Part of Ralph's cheerfulness and creativity may stem from the absence of mind-states of conflict and confusion that beset many of us!


Thanks, Dr Keeney, for answering my question (the last one, about game theory). Though you didn't label it as such, it sounds like you're saying one should choose the option with the best expected value -- which makes perfect sense.

Patrick Noonan

I think a couple of readers misunderstand. Not only does Ralph not "presume a degree of rationality to human behavior that just isn't so," but it actually is motivated in large part by that gap between what we want and what we do.

Rationality is often described as being mostly about consistency: Acting consistently with our preferences (our values, objectives - what we want) and our beliefs (about uncertainties - what we think will happen along each possible road we could take).

Our brains have a hard time sorting all that out, and Ralph's work (ditto the work of Howard Raiffa, who commented above, and who trained both Ralph and myself) is intended to provide practical therapy to narrow the gap.


Dr. Kenney:
Thanks for your Q&A session. I look forward to reading your book.


Great responses!!

Eric Vance

The way I suggest people make complex and important decisions such as "Should I move to Ireland to be with my girlfriend?" is to make a spreadsheet and then use my intuition.

For example, first I would list the various objectives I consider important "Proximity to girlfriend", "Salary", "Weather", "Future job opportunities". Second, I would list all possible decisions: "Move to Ireland", "Stay in London", "Live 50/50", or "Move to Afghanistan". Then pick the two, three, or four best options (cull the stupid options).

The next step is to assign point values to each decision for each of the objectives. How many points for the "Proximity" question does Option1 get? Option 2....

The second to last step is to assign point values for an "Intangibles" category. How many "intangible" points would the decision to move to Ireland be worth? These intangible points should be for the category "Everything else not on this spreadsheet."

The last step is to see which decision option has the highest intangible points and pick it. Ignore all the other points. You'll probably find that if you sum up all the categories, the option with the highest intangible points also has the highest sum for the other points.


data mining courses

His insights are great, and I really connect with how he describes the importance of decision making.


This is amazing work. Keeney's answers to the questions were eye openers. I learned GREAT deal.


Decision making is key to setting and achieving goals. They say successful people are the ones who are able to make good decisions efficiently and consistently.

Ulrike Wood-Sichra

I am borrowing a question from a dear friend of mine, who is not among us any more:
You are searching for your mate ... you meet the candidates ... when do you stop? How do you know you have met the optimal candidate?