Emerging Perspectives on Judgment and Decision Research

p.1"The overriding goal of this book is to provide a forum for fresh perspectives on decision making. The aim is to expose readers to a wide variety of promising perspectives for enhancing the scope of judgment and decision-making research."

A decision support system helps a client make a decision. Perhaps computer chess programs are really decision support systems that are left to run completely on their own.

p.39"Even among those who build, use and study them, there is some disagreement about what exactly should and should not be considered a decision support system. Nevertheless, a definition consistent with most conceptions is that a decision support system is a computer-based system, typically interactive, that is intended to support people's normal decision making activities"

...The entire process by which an expert system is built is commonly called knowledge engineering. Most expert systems are designed to function the way human experts behave when they serve as consultants. Conversations between experts and clients are a normal, essential element of consultations. Thus, expert systems typically contain one final feature:

If an expert is familiar with the problem that he is analyzing, then he can use his memory to recall similar situations. If the situation has not happened that often in his own personal experience, then he will have to use a rule-based system that is based either on current theory or on the advice of other experts who are more familiar with the particular type of problem. Experts will perform well in environments where the events are predictable.

p.144"Shanteau (1992) argued that expert performance is constrained by task characteristics: Experts tend to perform well in tasks that are static and have a high degree of predictability, and where feedback is readily available. Similarly, Vicente (1998) argued that expert performance is constrained by the inherent predictability of the environment: Experts working in environments that afford a high degree of predictability can exhibit a higher level of performance."

p.148"Whether an expert will use memory-based processing will depend on how familiar they are with the particular judgment situation. If they are highly familiar, then processing will be memory-based, but if the situation is completely novel, they will use rule-based judgment."

Perhaps research into how a chess master makes a move in a chess game has been complicated by the routine behavior carried out without conscious attention - policies that the chess master has developed over time. As this next quote explains, policies are decision routines that once occurred with conscious attention, but have become so routine that they have become encapsulated in unconscious activity that is just beyond conscious attention. This is possibly similar to asking an adult what steps are required to drive a car in heavy, fast-moving traffic - the policies developed by the adult have evolved over time and have become a fixed behavior just beyond conscious thought. The repertoire of tactical possibilities (changing lanes, passing, accelerating, braking) and positional moves (guessing that the left lane will eventually offer a faster path than the right lane, guessing that a truck up ahead will accelerate slowly up a hill and create an opening for a lane change, etc) is similar to the situation faced by the chess master at the chessboard. Our concern would then be with the subconscious policies of the master in examining a position and determining which move is best.

p.396-397"Highly practiced behaviors, for instance, become routinized until they can be carried out without conscious attention... Policies are decision routines that originally occurred through conscious attention and evaluation, but that over time have become part of a routine (e.g., brushing one's teeth or taking vitamins). This perspective reminds us that routine sequences of behavior are typically dependent on thoughtful decision making in their origins... Decisions that contribute to the establishment of behavioral routines can be thought of as personal policy decisions. Once a person makes a decision with policy implications for routine behaviors, the routine can be established and then carried out repeatedly without much additional decision making, unless something interferes with the routine or there is a change in the person's goals... It is our contention that personal policy decisions are of fundamental significance in people's lives but that research on decision making has paid very little attention to these types of choices as a class."

One cannot look at decisions without looking at the goals that motivate individuals - a decision is only as good as its ability to help us reach a goal, or to help us make progress in reaching the goal.

p.399"Because decisions guide behavior, the goals that motivate the decision process become of fundamental interest in understanding why decisions are made and in providing benchmarks for measuring the qualitative success of decisions."

p.423"To move successfully towards goals, it is essential that decision makers pay close attention to both temporal and situational contexts so that they can monitor their direction of movement with respect to any given goal across any period of time. Decision makers may also want to change goals or goal priorities as temporal and situational contexts change."

Knowledge itself appears to have no magical powers in the process of decision making. You need to know what to do with the knowledge or when or exactly how to apply it. The actual execution of steps towards the goal is just as important as the timely, relevant and actionable information which drives those steps.

p.627"Knowledge of relevant facts is clearly necessary for expertise. Someone who does not know the facts about a domain will be unable to make competent decisions. Yet knowledge alone is seldom sufficient to establish that someone has expertise. For example, knowledge of the rules of baseball is not enough to know how to play a particular position, or to pitch a curve ball, or to manage a team in the bottom of the ninth inning. Each of these skills requires more than knowing what the facts are. They also require an understanding of what to do with the facts."

It seems that our computer chess program must have the capability to discriminate between similar but not identical situations - that is, if we desire an understanding of the game on the level of an expert. We might, however, not care whether our computer program understands what is going on, so long as it wins the game.

p.627-628"Expertise lies in knowing when to follow the rules - and when not to... Weiss and Shanteau argued that discrimination is necessary for expertise... The ability to differentiate between similar, but not identical, cases is a hallmark of expert performance; top experts make distinctions that novices miss... consistency is necessary for expertise; someone who cannot repeat his or her judgment of a case is not behaving as an expert."

p.653"A person's ability to predict the world is completely determined by how well the world can be predicted from the available data, how consistently the person uses the available data, and how well the person understands the world."

To make better decisions, we should study the cases where smart people failed to choose wisely, even though they had access to all the information which would have let them do so. Expertise, in its most basic state, seems to be a collection of general rules for information gathering and categorizing, combined with a database of exceptions to these rules and the 'cues' which indicate when the exceptions to the general rules should be applied.

p.658"One of the most important questions for anyone interested in improving human judgment and decision making... is why smart people, who often have the requisite competence, reason poorly or make stupid choices."