Cognitive biases, cognitive heuristics, and memory processes can exert a major negative impact on judgment and decision-making strategies. Cognitive biases are preconceptions or beliefs that can negatively influence clinical judgment. Cognitive heuristics are simple rules that describe how clinicians, and other people, make judgments and treatment decisions. Reliance on cognitive heuristics can be efficient because they are simple and they allow us to make judgments and decisions quickly and with little effort, but they are fallible and can lead clinicians to fail to learn from their experiences. With regard to memory, it should be obvious that clinicians will not learn from their experiences when their memories of those experiences are incorrect.
Several cognitive biases and heuristics will be described. Confirmatory bias occurs when clinicians seek, attend to, and remember information that can support but not counter their hunches or hypotheses. When psychologists ask questions that can confirm but not refute their impressions of a client, they are unlikely to make good judgments and decisions and they are unlikely to learn from their experiences. Similarly, psychologists are unlikely to learn from experience if their memories are distorted to support their preconceptions. Empirical research indicates that confirmatory bias does occur when psychologists work with clients (Haverkamp, 1993; Lee, Barak, Uhlemann, & Patsula, 1995; Murdock, 1988; Strohmer, Shivy, & Chiodo, 1990).
Hindsight bias describes how individuals, including mental health professionals, generate explanations for events that have occurred. Psychologists are generally unaware that knowledge of an outcome influences the perceived likelihood of that outcome (Fischhoff, 1975). In other words, after an event has occurred, people are likely to believe that the event was bound to occur. Results on hindsight bias have been replicated across a range of judgment tasks (Hawkins & Hastie, 1990), including the diagnosis of neurological impairment (Arkes, Faust, Guilmette, & Hart, 1988). Hindsight bias is important for understanding why mental health professionals have difficulty learning from clinical experience because it suggests that they think in deterministic (not probabilistic) terms. As observed by Einhorn (1988):
The clinical approach to diagnosis and prediction can be characterized by its strong reliance on attempting to explain all the data. Indeed, a significant feature of diagnostic thinking is the remarkable speed and fluency that people have for generating explanations to explain any result. For example, "discussion sections" in journal articles are rarely at a loss to explain why the results did not come out as predicted (cf. Slovic & Fischhoff, 1977); psychotherapists are quick to point out that a patient's suicide should have been anticipated; and commissions, panels, committees, and the like, place blame on administrators for not knowing what is "obvious" in hindsight. As Fischhoff (1975) has pointed out, the past has few surprises but the future has many. (p. 63)
Mental health professionals will have trouble learning from experience if they do not recognize that all assessment information is fallible and that we frequently cannot make predictions with a high degree of certainty. That is, they will believe they have learned many things from a case when they have not. In conclusion, the cognitive processes described by the hindsight bias can lead clinicians to the erroneous belief that a particular combination of symptoms or behaviors is almost invariably associated with a particular outcome.
With regard to cognitive heuristics, the heuristic that is most relevant to understanding why clinicians can have a difficult time learning from experience is the availability heuristic (Kahneman, Slovic, & Tversky, 1982). This heuristic describes how selective memory can lead to judgmental error. Mental health professionals typically recall only selected information about a case because it is difficult, or even impossible, to remember all the details about a client. If their memories of a case are inadequate, they will have trouble learning from the case. According to the availability heuristic, the strength of a memory is related to the vividness of information and the strength of verbal associative connections between events. For example, a mental health professional is likely to remember a client who is striking or unusual in some way. Similarly, when trying to remember if a test indicator and a symptom or behavior co-occurred, a mental health professional may be influenced by the verbal associative connections between the test indicator and the symptom or behavior.
Finally, a large body of research on covariation misesti-mation suggests that mental health professionals are more likely to remember instances in which a test indicator and symptom are present than those in which a test indicator is absent and a symptom is either present or absent (Arkes, 1981; Kayne & Alloy, 1988). To learn whether a test indicator can be used to describe a symptom, one has to remember instances when the test indicator is absent as well as instances when it is present. Of course, an illusory correlation is said to be present when clinicians cannot accurately determine how two events covary. Thus, in the Chapman and Chapman (1967) study on illusory correlation, when undergraduates mistakenly remembered that there is a positive relation between unusually drawn eyes and watchfulness or suspicious-ness, they may have been remembering cases when clients drew unusual eyes but forgetting cases when this drawing characteristic was not present. To be more specific, if a significant proportion of clients who draw unusual eyes are watchful or suspicious, then clinicians may believe this is a valid indicator. However, if a significant proportion of clients who do not draw unusual eyes are also watchful or suspicious, then it would be inappropriate to conclude that unusual eyes is a valid indicator. Thus, covariation misestimation, in addition to verbal associative connections (as mentioned by Chapman & Chapman), may in part explain the occurrence of illusory correlation phenomena.
One other theory about memory and clinical judgment will be mentioned. The act of making a diagnosis can influence how a mental health professional remembers a client's symptoms (Arkes & Harkness, 1980). According to this theory, a mental health professional may forget that a client has a particular symptom because the symptom is not typical of the symptoms associated with the client's diagnosis. Similarly, a symptom that is typical of the diagnosis may be "recalled," even though the client may not have that symptom. Of course, it is difficult to learn from experience when the details of cases are remembered incorrectly.
Mental health professionals learn from experience when they receive unbiased feedback, but the benefits of feedback are likely to be setting specific. In several studies (Goldberg & Rorer, 1965 and Rorer & Slovic, 1966, cited in Goldberg, 1968; Graham, 1971), psychologists made diagnoses using MMPI profiles. They became more accurate when they were told whether their diagnoses were valid or invalid, but only when all of the MMPI protocols came from the same setting.
Unfortunately, mental health professionals typically do not receive accurate feedback on whether their judgments and decisions are valid. For example, after making a diagnosis, no one comes along and tells them whether the diagnosis is correct or incorrect. They sometimes receive helpful feedback from a client, but client feedback is subjective and can be misleading. In contrast, when physicians make judgments, they frequently receive accurate feedback from laboratory results, radiology studies, and, in some cases, autopsies. In most cases, for mental health professionals to determine the accuracy of a judgment or decision, longitudinal or outcome data would have to be collected. Longitudinal and outcome data are collected in empirical studies, but most clinicians find this data to be too expensive and time consuming to collect in clinical practice.
Client feedback can be misleading for several reasons. First, clients may be reluctant to dispute their therapists' hypotheses. This can occur if clients are passive, suggestible, fearful of authority, or motivated to be pleasing. Second, clients may be unable to give accurate feedback because they may not be able to describe all of their traits and symptoms accurately. Even their reports of whether they have improved will be subjective and will be influenced by how they feel when they are asked. Finally, mental health professionals may describe clients in general terms. Their descriptions may be true of clients in general and may not describe traits that are specific to a client (e.g., "You have a superb sense of humor" and "You have too strong a need for others to admire you"— from Logue, Sher, & Frensch, 1992, p. 228). This phenomenon has been labeled the Barnum effect, after the circus figure P. T. Barnum (Meehl, 1954). Occurrence of the Barnum effect will be misleading to clinicians if they believe their judgments and decisions are valid for a specific client and not for clients in general.
Client feedback will also be misleading if clinicians make incorrect interpretations but convince their clients that they are correct. For example, after being told by their therapists that they were abused, some clients falsely remember having been abused (Loftus, 1993; Ofshe & Watters, 1994). These therapists have used a variety of techniques to help clients believe they remember having been abused, including telling them that they were abused, repeatedly asking them to remember the events, interpreting their dreams, hypnotizing them, and referring them to incest-survivor groups. Of course, clinicians will have a hard time learning from experience if they convince clients to accept incorrect interpretations and judgments.
Was this article helpful?