Attention, Problem Solving and Intelligence (AAMC 6B)
Content category 6B covers how the brain makes sense of incoming information. The cognition, consciousness and language page introduces heuristics, Piaget, sleep and language, and learning and memory covers memory. This page adds the pieces that page only touches: attention, the full toolkit of problem solving and decision making, and intelligence, including how it is measured and how cognition changes with age. These topics are frequently tested through experiments, so the research designs are included throughout.
Attention
Attention is the process of concentrating limited mental resources on some information while ignoring the rest. It is needed because the senses deliver far more information than working memory can handle.
- Selective attention
- Focusing on one input while filtering out others, such as following one conversation in a crowded room.
- Divided attention
- Attending to two or more tasks at once. Performance usually drops unless at least one task is highly practiced (automatic).
- Sustained attention (vigilance)
- Maintaining focus over a long period, such as monitoring a radar screen. Performance declines with time on task.
Filter models of selective attention
| Model | Where unattended input is filtered | Key evidence |
|---|---|---|
| Broadbent's early selection | Before meaning is processed; only physical features (pitch, location) pass | In dichotic listening, people remember almost nothing of the unattended ear |
| Treisman's attenuation | Unattended input is turned down, not blocked; important items can still break through | People notice their own name in the unattended ear |
| Deutsch and Deutsch late selection | After all inputs are processed for meaning; selection happens just before response | Meaningful words in the unattended channel can influence responses |
The cocktail party effect, hearing your own name across a noisy room, is the standard challenge to strict early selection. In a dichotic listening task, participants hear two messages, one in each ear, and repeat aloud (shadow) one of them; researchers then test what was absorbed from the other ear.
Failures of attention
- Inattentional blindness: failing to notice a fully visible but unexpected object because attention is elsewhere, such as missing a person in a costume walking through a ball-passing game.
- Change blindness: failing to notice a change in a scene, such as a different conversation partner after a brief interruption.
- Attentional blink: after detecting one target in a rapid stream, a second target appearing within about half a second is often missed.
The Stroop effect shows that automatic processes can be hard to suppress. Naming the ink color of the word RED printed in blue is slower than naming the color of a row of Xs, because reading is so automatic that it interferes.
Problem Solving
A problem has an initial state, a goal state, and obstacles between them. Well-defined problems (solving an equation) have clear goals and paths; ill-defined problems (choosing a career) do not.
| Strategy | How it works | Trade-off |
|---|---|---|
| Trial and error | Try solutions until one works | Works for few options; inefficient for many |
| Algorithm | A step-by-step procedure that guarantees a solution | Guaranteed but can be slow |
| Heuristic | A mental shortcut or rule of thumb | Fast but can produce errors |
| Means-end analysis | Break the gap between current and goal state into subgoals and reduce each | Useful for complex problems; can miss detours |
| Working backward | Start from the goal and reason back to the start | Useful when the goal is clear and the start is not |
| Insight | Sudden restructuring of the problem (an aha moment) | Unpredictable; often follows an impasse |
Incubation, setting a problem aside, sometimes helps because it lets an unhelpful mental set fade.
- Mental set
- Persisting with an approach that worked before even when a simpler one is available.
- Functional fixedness
- Seeing an object only in terms of its usual function, such as not thinking of a coin as a screwdriver.
- Confirmation bias
- Seeking evidence that supports an existing belief and ignoring evidence against it.
Decision Making and Its Biases
The cognition page defines the availability and representativeness heuristics. Several further biases appear in MCAT passages, especially in medical decision scenarios.
- Framing effect
- The same choice is judged differently depending on whether it is described as a gain or a loss. A treatment with a 90% survival rate is preferred over one with a 10% mortality rate, although they are identical.
- Anchoring
- Estimates are pulled toward an initial number, even an irrelevant one.
- Overconfidence
- People are more certain of their judgments than their accuracy justifies.
- Sunk cost fallacy
- Continuing an investment because of what has already been spent rather than future value.
- Belief bias
- Judging an argument's logic by whether its conclusion is believable.
- Hindsight bias
- After an outcome is known, believing it was predictable all along.
Deductive reasoning moves from general premises to a specific conclusion that must be true if the premises are true. Inductive reasoning moves from specific observations to a probable general rule. Most scientific generalization is inductive, which is why findings are always provisional.
A study gives two groups the same information about a surgery. Group A reads that 85 of 100 patients are alive after five years; group B reads that 15 of 100 patients have died within five years. Group A is more likely to choose surgery.
The outcome information is identical, so the difference in choices can only come from the wording: this is the framing effect. Anchoring would require an initial number that biases a numerical estimate; availability would require ease of recalling examples. Because participants were randomly assigned to the two wordings, the study supports a causal claim about framing.
Intelligence
Intelligence is the ability to learn from experience, solve problems and adapt to new situations. Theories disagree mainly on whether it is one general ability or several.
| Theory | Core claim |
|---|---|
| Spearman's general intelligence (g) | A single general factor underlies performance on all mental tests, because scores on different tests correlate positively |
| Thurstone's primary mental abilities | Seven separate abilities (such as verbal comprehension, numerical ability, spatial ability) |
| Gardner's multiple intelligences | Eight or more relatively independent intelligences, including linguistic, logical-mathematical, spatial, musical, bodily-kinesthetic, interpersonal, intrapersonal and naturalistic |
| Sternberg's triarchic theory | Analytical (academic problem solving), creative (handling novelty) and practical (everyday street smarts) intelligence |
| Cattell and Horn | Fluid intelligence (reasoning about new problems) vs crystallized intelligence (accumulated knowledge and vocabulary) |
| Emotional intelligence | Perceiving, understanding, managing and using emotions |
Measuring intelligence
Binet created the first practical test to identify French children who needed extra help, using the idea of mental age. The original intelligence quotient was mental age divided by chronological age, multiplied by 100. Modern tests such as the Wechsler scales use a deviation IQ: scores are standardized to a normal distribution with a mean of 100 and a standard deviation of 15, so about 68% of people score between 85 and 115 and about 95% between 70 and 130.
- Standardization
- Giving a test to a large representative sample under uniform conditions to establish norms.
- Reliability
- Consistency. Test-retest reliability means similar scores on repeated testing; internal consistency means items agree with each other.
- Validity
- Measuring what the test claims to measure. Predictive validity means scores predict a later outcome, such as school grades. A test can be reliable without being valid, but not valid without being reliable.
The Flynn effect is the rise in average raw IQ scores across many countries during the twentieth century, about 3 points per decade. The change is too fast to be genetic, so it is attributed to environmental factors such as better nutrition, schooling and familiarity with abstract thinking. Tests are periodically renormed to keep the mean at 100.
Heritability says nothing about an individual (it is not true that half of one person's IQ comes from genes), and it does not explain differences between groups. When environments are made more equal, heritability goes up, because the remaining variation is more genetic. See behavioral genetics and development for twin and adoption designs.
Cognition Across the Lifespan
Piaget's stages describe how thinking develops in childhood (see the cognition page). In adulthood the main pattern is that fluid intelligence (processing speed, working memory, novel reasoning) peaks in early adulthood and declines gradually, while crystallized intelligence (vocabulary, general knowledge) is stable or rises into late adulthood.
Cross-sectional studies, which compare different age groups at one time, exaggerate decline because older cohorts had less education (a cohort effect). Longitudinal studies, which follow the same people over time, show smaller declines but suffer from attrition and practice effects.
- Normal aging
- Slower processing, more tip-of-the-tongue moments, mild retrieval difficulty; daily functioning intact.
- Dementia
- A decline in memory and other cognitive abilities severe enough to impair independent daily functioning. Alzheimer's disease is the most common cause, associated with amyloid plaques, tau tangles and loss of cholinergic neurons.
- Delirium
- Acute, fluctuating confusion and disturbed attention, often due to illness, medication or infection; usually reversible.
Cognitive abilities are also shaped by culture and environment. Exposure to schooling, enriched environments and continued mental and physical activity are associated with better cognitive outcomes in older age, though the direction of causation is debated.
Worked MCQs
5 MCQs in the style of the Psych/Soc section, several built around a study design.
Q1. In a dichotic listening task, participants shadow the message in their right ear. Afterward, most report hearing their own name in the left ear but cannot report any other left-ear content. This finding is most difficult to explain with which model?
Broadbent's model filters out unattended input before meaning is processed, so a meaningful item such as one's own name should not get through. Treisman's attenuation model and late selection models both allow highly meaningful unattended items to be noticed.
Q2. A test produces nearly identical scores when the same people take it twice, one month apart, but its scores do not predict later job performance. The test has:
Consistent scores over time indicate high test-retest reliability. Failure to predict the criterion it was meant to predict indicates low predictive validity. A test can be reliable without being valid.
Q3. A study finds that 70-year-olds score lower than 30-year-olds on a timed reasoning task but higher on a vocabulary test. The best interpretation is that:
Timed novel reasoning taps fluid intelligence, which declines gradually from early adulthood. Vocabulary taps crystallized intelligence, which is stable or rises with age. Because the study compared different age groups at one time, cohort effects could exaggerate the fluid difference.
Q4. Physicians are told that a new drug 'prevents 20% of strokes' or that 'stroke still occurs in 80% of patients who would otherwise have one.' More physicians prescribe the drug under the first wording. This demonstrates:
The two descriptions carry the same information but emphasize a gain versus a loss, and the choice shifts with the wording. That is the framing effect. Availability depends on how easily examples come to mind, which is not manipulated here.
Q5. Twin studies estimate that a trait has a heritability of 0.6 in a population. Which statement is justified?
Heritability describes the proportion of variance within a specific population associated with genetic differences. It does not apply to individuals, does not mean the trait is fixed, and cannot be used to explain differences between groups.
Quick Recap
- Selective, divided and sustained attention; dual-task studies reveal capacity limits.
- Early selection (Broadbent) vs attenuation (Treisman) vs late selection; the cocktail party effect challenges early selection.
- Inattentional blindness, change blindness and the attentional blink are failures of attention.
- Practice moves tasks from controlled to automatic; the Stroop effect shows automatic reading interfering.
- Algorithms guarantee a solution; heuristics are fast shortcuts; mental set and functional fixedness block solutions.
- Framing, anchoring, overconfidence, sunk cost, belief bias and hindsight bias distort decisions.
- Spearman's g vs Gardner's multiple intelligences vs Sternberg's triarchic theory; fluid vs crystallized intelligence.
- Deviation IQ: mean 100, SD 15. Reliable is not the same as valid. Heritability is about population variance, not individuals.