● Illusory relation

○ Unusual numbers = especially noticeable and more memorable

○ Mistakenly perceived to happen often
● Aysmmetric salience
○ Events vs non events (non-eventful, business as usual)
○ Surprising vs expected
○ Unusual vs common
○ Disruptive vs nondisruptive (dirty dishes vs clean dishes)
● Selective memory
○ Some events/experiences more memorable than others

● Broken door close button

○ Would it be immediately noticed?
○ How would people react if button doesn’t work?
○ You may have experienced this w/i being aware
■ Placebo buttons
● Elevator close
● Pedestrian crossing buttons
● Thermostats in offices
■ Why might we perceive placebo button has effect?
● Door eventually closes after pressing button → expected outcome
eventually occurs
● Inference: our action caused the closing
● ^example of “Illusion of control”
● Illusion of control
○ Door open → press button → door closed
○ Door open → no button press → door closed ← but may not be experienced or
noticed
○ Timing might be a cue

○
○ ^try to find explanations
■ Lack of immediate effect = potential cue BUT could perceived as delayed
effect
 ○ Why have placebo thermostats?
■ Perceived control → not possible/practical have individual settings over
conditioning systems (because of centralised AC)
■ Familiarity → feel more stressed when no control/button
○ Illusion of control
■ Coincidental r/s b/w actions and observable effects
■ Actions = often misinterpreted as CAUSING the effects
○ Examples
■ Placebo buttons
■ Gambling
■ Superstitious routines (e.g. sports)

● Origins of false beliefs

○ Neglect of coincidence
■ Fail notice outcomes may from chance
■ E.g. non-specific predictions, Barnum effect
○ Illusory correlations
■ Asymmetric evidence → perceive relationships that not actually present
○ Misinterpretation of random effects
■ Causal interpretation of random effects
● E.g. clustering illusion, regression fallacy

● Why persist?
○ Difficult change because of:
■ Neglect missing info
■ Confirmatory evidence = more salient
■ Tend seek confirmatory evidence
■ Biased interpretation of ambiguous info
■ Counterevidence = often explained away ← evidence = unambiguous,
obvious, noticed
● Reaction to contradictory evidence
○ Unlikely reactions:
■ Accept the evidence
■ Ignore evidence
■ Misperceive the evidence
○ Expectations = EXERT bias → BUT perception and memory = NOT radically
distorted
○ Tendency: apply scrutiny in selective way
■ Consistent evidence = accepted readily
■ Attempt to “ explain away” evidence that’s inconsistent w/ beliefs
■ E.g. tongue map
● “Explaining away” could = persistent of tongue map myth
● Inconsistent evidence could explained away by:
○ Error w/ test
 ○ Not localised
○ Individual differences
^seek alternative explanation that wouldn’t require revision of what
is already believed
● Why do gamblers persist?
○ Gilovich (1983): conversations of gamblers ab wins/losses
○ Do gamblers remember and focus on wins more? Memory or attention bias?
■ NO ← not the case
■ BUT systematic differences in how wins/losses interpreted
● Wins
○ Outcome perceived as appro and predictable
○ Discount random elements
● Losses
○ Interpreted as “flukes” due to atypical factors
○ Emphasise how could’ve potentially different
→ accepted wins BUT explained away losses
● Death penalty evidence
○ Lord et al (1979)
■ 2 groups → for and against death penalty
■ Presented w/ info that was carefully balanced → some supporting, some
against
■ Outcome
● Beliefs tend REINFORCED for ALL subjects ← regardless
originally for/against

^critique contradictory evidence can = reinforce belief

^tend more cautious/critical when faced w/ contradictory evidence
● Avoiding bias
○ Strategy: consider possibility of coincidence
■ Tend seek causal interpretations, and neglect possibility of coincidence
■ Keep this in mind, and recognise specific situations
 ■ E.g. regression to mean
○ Strategy: be aware of missing info
■ Info required to evaluate belief = often unavailable or unoticeable
■ Ask yourself: how would I know if true or not? Is the info even available?
■ Notice situations can cue you to be skeptical about beliefs
■ Strategy: be aware of potential forbias
● Accept ALL susceptible to biases due to expectations and prior
beliefs
● Can look for situations where bias is likey, and apply MORE
skepticism
○ E.g. gender stereotypes
○ Should cautious about accepting generalisations a/b men
vs women
■ Strategy: study science
● Persistence of false beliefs = pose challenge for scientific advance
● Scientific culture and methods = developed to overcome cognitive
biases
○ Scientific method: testing falsifiable predictions
○ Appreciation of randomness, use of statistics
○ Peer-review, competitive marketplace

● Ineffective health remedies

○ Many traditional “treatments” were ineffective or dangerous
○ E.g. bloodletting, mercury
○ How did these “treatments” persist for years w/o being
noticed as being ineffective?
○ One factor: will to believe
■ Patients = in vulnerable situation and open to possibility of hope
■ Tend open to trying novel, unusual treatments

○ Spontaneous improvement
■ Many conditions get better on their own
■ Estimate: 50% of conditions that causes us seek medical help actually
improve w/o treatment
■ Very likely condition will improve REGARDLESS of treatment
■ BUT if receive treatment and get better, tend seek explanation for this
● Very likely to mistakenly attribute improvement to treatment (NOT
self-healing)

○ Spontaneous fluctuations
■ Even if conditions get worse, probably RANDOM fluctuations over time
■ Likely seek medical help in BAD state → so likely improve or stop getting
worse after getting medical help (regardless of treatment)
 ■ High potential for illusory causation

○ Interpretation of failure
■ E.g. treatment = NOT effective at curing cancer
■ Even if treatment has merit, wouldn’t expected cure every time
■ Attribute failure to PRACTITIONER rather than method
● E.g. treatment NOT administered properly, dosage amount
inappropriate
■ Attribute failure to PATIENT
● Patient waited too long to seek treatment
● Something atypical a/b patient
○ Non-specific criterion
■ If expected result = NOT clearly specified → prevent evaulation of
practice

● Factors in belief of Ineffective health remedies

○ Spontaneous recovery and fluctuations
■ High likelihood of spurious/false correlation → tend
perceived as causal
○ Easy to discount -ve evidence
■ Failures could = attributed to:
● Practitioner ← didn’t carry out procedure correctly
● Patient ← e.g. too late to get treatment
■ Non-specific criterion for success
● Not clearly define expected benefit → hard
disprove effectiveness

● Test a possible treatment

○ Intuition NOT trustworthy
■ Limited amount of direct experience
■ Tendency see patterns even if none
■ Interpretation of evidence = easily biased
○ Controlled experiment needed to test whether it REALLY works
● Scientific approach
○ Modern standard: controlled clinical trials
■ Treatment vs control group
● Controls for possibility of spontaneous recovery
● Concrete evidence can be found to prove/distinguish

● Control group may receive placebo

○ Controls for bias due to patients’ expectations

● Double-blind → researcher doesn’t know who’s receiving treatment

○ Control fo rbaises due to researchers’ expectations

● Statistical analysis of effects

○ Distinguishes real effects from random
● Predetermined method and sample size
○ Avoid adjusting method and adding subjects until evidence for success
● Predetermined, objective criteria
○ Avoid search confirmatory evidence, using non-specific criteria
^try to help overcome biases

● Are scientists biased?

○ Biases = general tendencies
○ Misinterpretation of evidence = side effect of mental processing that allows us
adaptively use prior knowledge
● Pseudo-science
○ E.g. carniology
 - Adjust info to avoid disrupting a belief that’s fundamental

● Failed replications in psychology

○ E.g. Ego depletion → less willpower after one exerts willpower
○ E.g. physically smiling can = happier
○ E.g. Power pose → more willing accept risk, higher testosterone
○ E.g. Elderly-slow priming → walk slower after age-related words
^famous psychological effects have failed to be replicated

● If effects aren’t real, what might’ve caused them?

○ Selective reporting of results
■ Report results that confirm predictions, NOT conflicting findings
○ Selective choice of analyses
■ Choose analyses that yield “right” results
○ Biased interpretation
■ Try “explain away” conflicting findings, BUT neglect potential
problems for confirmatory

● How scientific practices can help counteract biases:

○ Pre-registration of methods and analysis plans
○ Encouraging replication studies in addition to novel research
○ Public sharing of data and materials

● How doctors think

 ○ Training: deliberate process of analysing clinical info

○ Estimated time b/w meeting patient and forming opinion = 20 sec

○ Didactic (taught/trained) approach used in training and case reviews: 20-30 mins
■ Doctors = use heuristics or pattern recognition

● Case study: Anne Dodge

○ Whenever referred to new specialist, also receive case history → specialists NOT
starting from neutral state → anchoring to original diagnosis
○ BUT evidence inconsistent w/ anoerxia/bulimia diagnosis
■ Put on 3000 cal/day diet BUT still losing weight
■ BUT biased interpretation of this evidence → assumed be LYING
(“explained away”)
○ Actual diagnosis: celiac disease (essentially a gluten allergy)
○ SO high calorie diet w/ lots of cereal and pasta = made it worse
○ Specialist that correctly diagnosed DIDN’T focus on abdomen/irritable bowel
○ Looked from fresh perspective, not just taking the psych evaluation at face value

● Causes of medical errors

○ Cognitive biases
■ Availability
■ Anchoring
■ Confirmation bias
●
○ Doctor had knowledge of aspirin toxicity but WHY was this originally missed?
■ Initial bias: availability
● Many recent cases of pneumonia → easily available diagnosis
● Confirmation bias caused error to persist
○ Some evidence consistent w/ pneumonia, others
inconsistent
○ Doctor discounted inconsistent evidence while accept
consistent evidence