In Parts 1 to 4, we traced the structure of overconfidence from four angles: why overconfidence is called the biggest trap, metacognitive blind spots, unconscious defense mechanisms, and self-serving attribution bias. These were the "Why Overconfidence Happens" chapters. From Part 5 onwards, we will move on to the chapter "How to notice". The central proposition of this issue is:Overconfidence occurs unconsciously, but its traces can be observed from the outside. That's what it means.language signs (shading of confidence expressions),signs of action (avoidance of verification and quick decision),signs of recognition (一致情報への偏重)、social signs (response to opposing opinions),body signs (Posture and Breathing) - Based on the five points of observation, Joshua Clayman et al.'s confidence interval research, Russell & Shawmaker's judgment research, Philip Tetlock's confidence-accuracy research, Edgar Schein's humble questions, Ekman's microexpression research, Xun Zi's Explanation, and Lao Tzu's Self-Knowledger Ming - Build your observation skills by layering seven points of view.
01Why is the skill to read "signs" necessary?
Overconfidence occurs unconsciously. This was the structure we followed up to the third session. If overconfidence is difficult to eliminate through conscious introspection, what is needed instead?Techniques for observing the traces left by overconfidence from the outside is. Traces, whether our own or those of others, appear in observable ways. Training to see this is the subject of this episode.
There are two premises for this article. Firstly,Overconfidence is not hidden. It is "hard to see", not "invisible". Traces remain in the form of language, actions, bodies, and social reactions. Second,Observation can be learned if you have the tools.. By converting signs into vocabulary, you can see things that you didn't notice before. This is of the same nature as the experiential knowledge accumulated over many years in medical diagnosis, quality control, and criminal investigation.
"Overconfidence in expert judgment is often not hidden, but simply hidden because the observer lacks vocabulary. Once we have a word to describe the signal of overconfidence, traces that we had previously overlooked suddenly come to the foreground." - J. Edward Russo & Paul J. H. Schoemaker, "Decision Traps" (1989) Purpose
本回が提供するのは、まさにその語彙です。五つの観察軸 ── 言語・行動・認知・社会・身体 ── それぞれに、過信が残す特徴的なパターンがあります。これらを意識的に観察する訓練が、自分の過信に対する早期警報装置になり、同時に他者の過信を理解する非難なき観察力になります。
02Signs of language ─ shading of expressions of confidence
Overconfidence manifests itself most clearly when we verbalize our judgments. US judgment researchers joshua clayman and jack sole showed in their 1999 paper "Overconfidence: It depends on how, what, and whom you ask" that how people express their confidence is directly linked to the core of overconfidence. The key to their discovery confidence interval The width of
"If you ask someone to give a range of answers that you are 90% sure about, the probability of a correct answer falling within that range is actually only 30-60%. People unconsciously set their confidence intervals much narrower than they actually are. Expressions such as 'definitely,' 'definitely,' and 'definitely will be' are the most typical linguistic traces of overprecision." -- Klayman, Soll, Purpose from González-Vallejo & Barlas (1999)
Signs of language are adverbs and modifiers expressing confidence. Density Appears on. When you have a vocabulary of observation, these things will show through, both in what you say and what others say.
The strongest expression of confidence
“Absolutely,” “without a doubt,” “certainly,” “100%,” and “without a doubt.” In light of Klayman et al.'s experimental data, the more this expression appears in a judgment, the more the actual probability tends to be lower than the expression suggests.
strong conviction expression
"Obviously", "Of course", "Of course", "Needless to say". A signal that is convinced not logically but intuitively. Words that keep away opportunities for validation.
expression of exclusion
"It can't be..." "There's no way..." "I can't think of it" "That can't happen." Linguistically eliminates the very possibility. Linguistic manifestations of the defense mechanism of denial (part 3).
When these expressions appear, it doesn't necessarily mean that we are really sure. rather There is a psychological need to strongly express one's convictions. Very likely. The first etiquette in this article is to practice internally asking yourself, what is the actual confidence interval for the judgment when you or someone else uses these expressions in the field of material evaluation?
03Behavioral Signs - Verification avoidance and hasty decision making
Overconfidence also leaves traces in your actions.J. Edward Russo and paul showmaker 's 1989 business decision study, Decision Traps, described three behavioral patterns common to overconfident decision makers.
speed of judgment
The time required to consider information is extremely short in areas where there is similar experience. Make a quick decision with a feeling of "I've seen this before." Although the complexity is not uniform, the decision time is uniform. Kahneman's overreliance on System 1.
Avoid searching for counter-evidence information
They eagerly collect information that supports their judgments, but are not proactive in searching for information that contradicts their judgments. I don't have a good habit of "checking just to be sure." Behavioral manifestations of confirmation bias (Wason 1960).
Omission of review
After making a decision, there is little habit of reflecting on how that decision turned out in hindsight. We look back on the good times and don't look back on the bad times, blaming them on the situation (Self-serving bias #4). Systematic loss of learning opportunities.
These three behavioral patterns are efficient judgment style It is experienced as. Fast, decisive, and unwavering -- these are qualities that are often admired. That's why observers need to look for the shadow of overconfidence in those same traits.A quick decision and a decision are superficially similar, but the latter is a conclusion based on searching for contrary information, while the former omits it.. The practice of this section is to train you to discern the differences.
04Signs of cognition: Overemphasis on matching information
Signs at the cognitive level behind behavioral bias Confirmation Bias is. british psychologist peter wason (1924–2003) demonstrated this phenomenon in his 1960 paper ``On the failure to eliminate hypotheses in a conceptual task,'' which became one of the central concepts in judgment research for the next half century.
"When testing a hypothesis, people preferentially gather information that is consistent with the hypothesis, and show systematic resistance to gathering information that contradicts the hypothesis. Logically, even though the search to disprove a hypothesis yields more information, people tend to choose confirmation over falsification." - Peter Wason, Quarterly Journal of Experimental Psychology (1960) Purpose
Confirmation bias leaves traces that can be observed from the outside. When a judge talks about information, how much information that is consistent with that judgment is foregrounded, and how much contradictory information is mentioned -- just looking at the ratio reveals cognitive bias.
Selective enumeration of matching information
When discussing a judgment, examples, data, and experiences that support the judgment come up one after another, but contradictory cases are omitted or treated as "exceptions." The most easily observed surface of confirmation bias.
Immediate rebuttal to counterargument
When an opposing opinion is raised, a counterargument is immediately assembled without taking time to listen. There is no time to internally consider the logic of opposing opinions. "Rebut before you listen" structure.
Making an exception an "exception"
When you come across a case that doesn't fit your judgment framework, you don't use it as material to reconsider your judgment framework, but instead treat it as an exception as ``this is a special case.'' An everyday version of what Lakatos calls an "ad hoc rescue."
These three signs are important to the individual. talking logically It is experienced through the senses. Therefore, observation requires skill. As a practice in this section, when talking about judgment,その人が "矛盾する情報" を何回口にしたかをカウントする Suggest training. If you keep getting zero times, it's a strong sign of confirmation bias. This can be used both to observe others and to reflect on yourself.
05Social Signals -- Reactions to Opposition
Overconfidence leaves its mark even within a group. Mentioned in Part 1 Irving Janis Groupthink research has shown that overconfidence within organizations How to handle opposing opinions showed that it appears most clearly in
Organizations with strong groupthink have several observable symptoms. There is an atmosphere in which it is difficult to raise opposing opinions at decision-making meetings. Those who express minority opinions are criticized as ``opposing for the sake of opposing'' and ``breaking the discussion.'' Early in the meeting, there is a sense that the direction is set. These are also organizational manifestations of Klein's projective identification, which I dealt with in Part 3.
"There are clear external signs of an organization in which groupthink is functioning: overreaction to dissent, self-censorship, disregard for outside information, a sense of superiority over 'us,' and underestimation of risk. When these things come together, the group's judgment is often worse than that of its individual members." - Irving Janis, Groupthink (2nd ed., 1982)
反対意見への即時防御
The moment an opposing opinion is raised, before we can consider it logically, we react with refutation or contempt. "That point of view has already been discussed." "That's a minority view." of opposing opinion logic よりも position is processed.
emergence of silence
Silence among members after the discussion took on an atmosphere of ``the conclusion is already in sight.'' Although it appears to be an agreement on the surface, there is a possibility that a third repression defense mechanism is operating throughout the group.
Disrespect for external information
Dismiss information and experience from other teams, departments, and organizations by saying, "Our case is different." Kelley's typical pattern that excludes discriminability information (see Part 4).
Social cues can be observed both in our own judgments and in the judgments of others. At meetings, how do you listen to whose opposing opinions? How are the opposing opinions I raise treated? Even if you don't leave a record, there are traces that you can notice when you look back.
06Body signs -- posture, breathing, gaze
Overconfidence also leaves traces on the body. American psychologist Paul Ekman (1934–)'s research on emotions, particularly microexpressions, systematized short-term emotional expressions that people cannot consciously control. Overconfidence leaves a contradiction between the willful feeling of avoiding validation and the slightest doubt in the form of subtle movements in the body.
However, physical signs should be treated with caution. As Ekman repeatedly warned,Inferring emotion or cognitive state from a single physical sign is unreliable. The combination of multiple signs and understanding the context support the accuracy of observations.
Fixed posture
His posture does not change while he is talking about his judgment. Your line of sight is fixed, neither leaning forward nor backward. The physical performance of "I'm sure" is unconsciously strengthened. Originally, if your thoughts are moving, your body will also move in subtle ways.
shallow breathing
My breathing becomes shallow while listening to opposing opinions. This is a physical sign that you are "pretending to listen and preparing a counterargument." The interval between a deep inhalation and a beat is a physical sign that new information is being processed internally.
Avoiding eye contact
Don't look into the eyes of the person you disagree with, or look too much into the eyes (try to read the other person's emotions). Both are signs that you don't have the time to internally consider the other person's logic.
Physical signs can also be used as clues to monitor your own condition. By paying attention to whether your own breathing becomes shallow or whether your posture becomes stiff as you make decisions, you may be able to detect the moment when a switch of overconfidence is turned on. The physical techniques from the previous series, ``To Gain Confidence,'' and the physical observations in this episode are a pair.
07Reading other people's overconfidence -- a study of Tetlock's confidence and accuracy
The empirical basis of the technique for observing overconfidence in others was mentioned in Part 1. philip tetlock in a long-term study. The most practical part of his discovery was"Overconfident forecasters can be identified by their expression of confidence." That's what it means.
"After 20 years of research, we have found that there is a weak negative correlation between the expression of confidence and the accuracy of predictions. Experts who predict with confidence that ``it will definitely happen'' are more likely to miss. Conversely, experts who say ``we should consider several scenarios'' and ``the probability is about '' were more accurate over the long term.'' - From Philip Tetlock's Expert Political Judgment (2005)
In 2015's Superforecasting, Tetlock describes a forecaster who is not overly confident. super predictor ── We have organized the linguistic features of. The language they use contrasts with that of overconfident forecasters.
Assertive/binary
“It will be,” “it will not be,” “never,” “never.” Reducing possibilities to binary values. There is no middle ground. probability Expression close to 0% or 100% Tendency to talk about it.
Probabilistic/gradual
"The probability of..." "If these conditions are met..." "If..." Be conscious of probability middle range A tendency to express conditions clearly.
The implications are direct for material reviewers.Just by observing how colleagues and superiors talk about their judgments in terms of verbal expressions of confidence, we can fairly infer the degree of overconfidence.. This is not an observation to criticize the judge, but an observation to correct the confidence interval of the judgment by oneself. It may not be appropriate to perform the same external risk assessment for a case that is said to be ``absolutely okay'' and a case that is said to be ``there is a 30-40% chance that a problem may occur.''
08Xunzi and Laozi ── Etiquette of self-knowledge
The technique of observing overconfidence in oneself and others is not something that modern psychology discovered. Ancient Chinese thinkers had already observed the same place more than two thousand years ago.Xunzi The story of Xunzi (circa 3rd century B.C.) is based on human judgment. What is it covered by (= becomes invisible)? This is a book that systematizes the
“Everyone's suffering is hidden by a single song, and the darkness does not come to Dali.
What Xunzi shows is that human judgment is always a crooked perspective It is a universal proposition that one can turn one's attention away from the whole reason. The ``ways of being hidden'' that he enumerated in his unraveling chapter can be said to be the forerunners of modern cognitive bias research: being hidden by desire, being hidden by aversion, being hidden by the beginning, being hidden by the end, being hidden by distance, being hidden by proximity, being hidden by breadth, being hidden by shallowness, being hidden by the past, and being hidden by the present. These are all cognitive distortions that have been rediscovered in modern judgment research.
Xunzi's prescription is“If all things are different, that is, they do not contradict each other.” ── Things always have different points of view, so if you stick to one point of view, it will always be obscured. As a manner of observation,Habit of reviewing one's decisions from an opposing perspective stand up. This is an ancient recipe for confirmation bias.
Chapter 33 of Lao Tzu's Tao Te Ching takes a verse that has been quoted several times in this series, but once again points to it from a different angle.
“Those who know others are wisdom, and those who know themselves are ming (knowledge of knowledge, ming of self-knowledge)” - Lao Tzu, Tao Te Ching, Chapter 33
Rereading it in the context of this episode, Laozi's "wisdom of acquaintances" is Techniques for reading signs of overconfidence in others, "self-knowledge person" is Techniques for reading signs of overconfidence is. The two are different abilities, and the latter is more difficult to master. This is consistent with Tetlock's finding that it is relatively easy to evaluate the predictions of others, but it is much more difficult to accurately estimate the accuracy of one's own predictions. The observation diary, which will be discussed in the final section of this article, is a practice for honing this direction of ``light''.
09At the pharmaceutical site ── 4 scenes of observation
Bringing abstract theory to the field. In the daily life of a material examiner, we will identify four specific situations where five signs (verbal, behavioral, cognitive, social, and physical) can be observed simultaneously. Each can be applied to either oneself or others as the object of observation.
The moment a decision is made at a review meeting
Observe statements that lead to conclusions. Concentration of belief expression (language), short judgment time (behavior), immediate reaction to counterarguments (social), fixed posture (body). When these four factors are present at the same time, the confidence interval for a judgment is likely to be wider than the surface belief.
The moment when a difficult matter is decided immediately
Observe quick decisions in areas with similar experience. Saying ``I've seen this before'', omitting confirmation, and the absence of contradictory information -- when all three are present, confirmation bias is turned on. A moment worth setting aside time for.
Immediately after objections were raised
Observe the length of the "pause" after an opposing opinion is raised. An immediate rebuttal is a sign of defensiveness, a beat of silence is a sign of internal deliberation, and a deep nod followed by a question is a sign of mature defensiveness (3rd Vaillant). The quality of the relationship indicates the degree of overconfidence.
Review of projects with good results
Observe the way students talk about their reflections. If you immediately answer, ``It's because of my judgment,'' then you are at the beginning of Langer's illusion of control (Part 4), and if you say, ``Several elements just happened to fit together,'' you are at the threshold of self-observation. The way a person speaks shows the depth of his or her knowledge.
これらの場面は、観察のための実験室です。自分自身を観察対象とすると、判断の最中は気づきにくいですが、振り返りの段階では痕跡が残っています。記録を残すと、観察の精度は飛躍的に上がります。次節がその実践です。
10How to write an observation diary ─ 4 questions
The art of reading signs of overconfidence is observation. habit It will be learned most efficiently if you cultivate it as a child. As a final practice in this article, we will propose a method for recording reflections on decisions in an observation diary. Each question takes about 3 minutes and involves answering four questions.
“What words did I use to express this judgment?”
The moment you make a decision, write down the expression of conviction you used. Did you use "absolutely" or "definitely" or "with a probability of..." or "depending on conditions"? As Klayman et al.'s study shows, linguistic expressions distort confidence intervals for judgments.
“What ratio did I consider between information that supports this judgment and information that contradicts it?”
Write down three examples/data that can be cited as the basis for your decision. Next, write down three examples/data that contradict your judgment. If the latter is zero, it's likely that Wason's confirmation bias was turned on.
“How did you hear the opposing opinions?”
Were there any opposing opinions at the meeting? How did you react if it happened? Did you immediately refute it? Did you take a breather? Did you take time to internally consider it? Self-observation of social signals.
“What is the current confidence interval for this judgment?”
Use your inner senses to estimate the probability that your decision is correct -- not a single probability, but something like ``50-80%.'' Training to apply the probabilistic expressions learned from Tetlock's super forecaster's language to one's own decisions.
The effects of making the four questions a habit will appear in two stages. In the short term, your self-evaluation of past decisions changes. The resolution of self-awareness increases from ``I made perfect decisions'' to ``There were some moments when my confirmation bias was turned on.'' In the long term, these questions will automatically come to mind while making decisions, and the quality of the decisions themselves will change. This is a method that also functions as a preliminary step to the calibration training that will be covered in Part 8.
Overconfidence occurs unconsciously, but its traces can be observed from the outside - this was the central proposition of this episode. Linguistic signs (Klayman et al.'s confidence interval research), behavioral signs (Russo & Schoemaker's judgment research), cognitive signs (Wason's confirmation bias), social signs (Janis' groupthink), physical signs (Ekman's facial expressions), signs of overconfidence in others (Tetlock's confidence-accuracy research), Xunzi's clarification and Laozi's self-knowledge ── Seven perspectives in different words Overconfidence isn't hidden, it's just a lack of vocabulary for observation. It was pointing to the same place.
The previous four chapters were about "Why overconfidence occurs." From this issue onwards, we will move on to the "how to notice" chapter. In the next installment (part 6), we will expand the scope of our observations to include the time axis. Why you can't learn from failure. Why do we not learn from past judgment errors that lead to updating our judgment frameworks? Tracing hindsight bias, organizational learning difficulties, and Argyris's single/double-loop learning, we dissect the structure by which overconfidence impedes learning.
- Overconfidence is not hidden. Traces appear in five axes: language (density of expressions of belief), behavior (avoidance of verification and quick decisions), cognition (overemphasis on consistent information), social reactions (handling of opposing opinions), and body (posture and breathing). If you have the vocabulary, you can observe it.
- Tetlock's discovery: high confidence has a weak negative correlation with prediction accuracy. In the long run, people who say ``the probability is about '' are more accurate than people who say ``it's definitely ''. Just by observing the verbal expression of confidence, you can correct the confidence interval of your judgment.
- The habit of using the four questions in your observation diary -- language, ratio of contradictory information, reaction to opposing opinions, and self-estimation of confidence intervals -- for three minutes to reflect on your decisions can become an early warning system for overconfidence. This also serves as a prelude to the 8th calibration exercise.
- Klayman, Joshua, Soll, Jack B., González-Vallejo, Claudia & Barlas, Sema. "Overconfidence: It depends on how, what, and whom you ask." Organizational Behavior and Human Decision Processes 79 (3), 1999, pp. 216–247. (Systematizing the relationship between confidence interval expression and overconfidence)
- Lichtenstein, Sarah, Fischhoff, Baruch & Phillips, Lawrence D. "Calibration of probabilities: The state of the art to 1980." In Kahneman, Slovic & Tversky (eds.), Judgment under Uncertainty: Heuristics and Biases. Cambridge: Cambridge University Press, 1982, pp. 306–334. (A classic in calibration research)
- Russo, J. Edward & Schoemaker, Paul J. H. Decision Traps: Ten Barriers to Brilliant Decision-MakingNew York: Doubleday, 1989. (Systematization of behavioral signs of overconfidence in business decisions)
- Russo, J. Edward & Schoemaker, Paul J. H. "Managing overconfidence." Sloan Management Review 33 (2), 1992, pp. 7–17.
- Wason, Peter C. "On the failure to eliminate hypotheses in a conceptual task." Quarterly Journal of Experimental Psychology 12 (3), 1960, pp. 129–140. (Classical experiment on confirmation bias)
- Nickerson, Raymond S. "Confirmation bias: A ubiquitous phenomenon in many guises." Review of General Psychology 2 (2), 1998, pp. 175–220. (Comprehensive review of confirmation bias)
- Janis, Irving L. Groupthink: Psychological Studies of Policy Decisions and Fiascoes. 2nd edition. Boston: Houghton Mifflin, 1982. (System of social signs of groupthink)
- Ekman, Paul. Emotions Revealed: Recognizing Faces and Feelings to Improve Communication and Emotional Life. 2nd edition. New York: Times Books, 2007. (Science of facial expressions and microexpressions. Japanese translation: Yasuhiko Suga translation, "The face lies as much as the mouth" Kawade Shobo Shinsha, 2006)
- Tetlock, Philip E. Expert Political Judgment: How Good Is It? How Can We Know? Princeton: Princeton University Press, 2005. (Relationship between confidence and accuracy)
- Tetlock, Philip E. & Gardner, Dan. Superforecasting: The Art and Science of PredictionNew York: Crown, 2015. (Linguistic features of super predictors. Japanese translation: “Super Predictive Power” translated by Nami Hijikata, Hayakawa Shobo, 2016)
- Schein, Edgar H. Humble Inquiry: The Gentle Art of Asking Instead of TellingSan Francisco: Berrett-Koehler, 2013. (How to respond maturely to opposing opinions. Japanese translation: The Art of Questioning, translated by Mayumi Kanai, Eiji Publishing, 2014)
- Xunzi's "Xunzi" (China, circa 3rd century BC). Explanation section.
- Laozi's Tao Te Ching (China, circa 6th century BC). Chapter 33, "Wisdom of the Acquainted, Ming of the Knower".