Cognitive approach

A mental framework of knowledge, built from experience, that organises information and guides expectations.

The cognitive approach — it explains behaviour through mental processes.

We attend more to schema-consistent information and may ignore what doesn't fit.

Recall reconstructs the event using the schema, adding expected details and dropping inconsistent ones.

Recalling books in an office that were never there, because the 'office' schema expects them.

Strong experimental support and wide application (memory, stereotypes, education).

Schemas can't be measured directly and the theory is vague on how they form or change.

Bias — memory is reconstructed, so expectations distort it.

A reconstruction — we rebuild events using schemas, not replay a recording.

Schemas can add vivid, expected details that feel real but never happened.

Learning by linking two things, so a once-neutral cue comes to trigger a response on its own.

A cue that causes no special reaction at first, like a beep before it means anything.

Something that triggers a reflex naturally, with no learning needed, like food.

The natural reflex to the UCS, like the mouth watering at food.

The once-neutral cue that now triggers the response after pairing, like the learned beep.

The learned reaction to the CS, like the mouth watering at the beep alone.

The stage where the cue and the natural trigger are paired again and again until the link is learned.

When the CS keeps appearing without the UCS, so the learned response slowly fades.

A faded learned response returning after a rest, even after extinction.

Reacting to cues that are similar to the conditioned stimulus, like a different beep.

Ivan Pavlov, who noticed his dogs drooled at a signal that came before their food.

Explaining learned fears (phobias): a harmless thing paired with a scary event can become frightening.

Learning in which the consequences of a behaviour make it more or less likely to be repeated.

Adding something pleasant after a behaviour, which increases it.

Removing something unpleasant after a behaviour, which increases it.

Adding something unpleasant after a behaviour, which decreases it.

Removing something pleasant after a behaviour, which decreases it.

Reinforcement increases behaviour; punishment decreases it.

Punishment suppresses behaviour but doesn't teach a better one, and can cause fear or resentment.

Operant = learning from consequences of behaviour; classical = linking two stimuli.

It ignores thoughts, emotions and social meaning, and much evidence is from animals.

Causality — a consequence causes future behaviour to change.

A simplified representation of how a mental process works, used to describe, predict and test it.

Because mental processes can't be seen directly; a model makes them concrete and testable.

Sensory memory, short-term memory, and long-term memory.

Attention — we transfer what we attend to.

Rehearsal (repetition).

By usefulness — whether they predict findings and guide research — not only by being fully true.

They turn invisible processes into testable predictions and generate experiments.

They oversimplify and can treat the mind too much like a computer.

No — the boxes and arrows are a useful description, not real physical parts.

Measurement — a model makes an invisible process concrete enough to study.

The idea that thinking uses two systems — fast, automatic System 1 and slow, effortful System 2.

Fast, automatic, effortless thinking — instant judgements and routine tasks.

Slow, deliberate, effortful thinking — logic, reasoning and new problems.

System 1 shortcuts that answer before System 2 checks.

The bat-and-ball puzzle: System 1 says '10p', but the answer is 5p.

System 1 shortcuts are usually helpful and fast; they only mislead in tricky situations.

It explains a huge range of biases and is supported by reasoning and reaction-time studies.

The 'two systems' may be a metaphor, not two literal parts of the brain.

No — it is fast and usually right; the skill is knowing when to slow down.

Bias — System 1 shortcuts produce systematic errors.

The tendency to seek, notice and remember information that fits existing beliefs and ignore what contradicts them.

Biased search, biased attention, and biased memory.

Believing a supplement works by noticing good days and forgetting bad ones.

Actively seek disconfirming evidence — ask 'what would prove this false?'

No — the filtering is mostly automatic (System 1), so it's hard to notice.

People seek sources that agree with them and dismiss those that don't.

Robust across many studies and explains stereotypes, poor decisions and echo chambers.

It describes a tendency that varies, more than it explains the exact cause.

Comparing belief vs no-belief conditions stops us counting only confirming cases.

Bias — expectations distort which evidence we use.

Relying too heavily on the first piece of information (the anchor) when making a judgement.

The first value or piece of information that becomes a reference point for a judgement.

We adjust from it but not far enough, so the answer stays pulled towards it.

A high 'original' price makes a sale price feel like a bargain.

Yes — even numbers known to be random still shift people's estimates.

A high opening offer anchors the final price higher (and vice versa).

Replicated in many experiments, even with random anchors.

Effect size varies and it describes the pattern more than the exact mechanism.

Form your own estimate before seeing anyone else's number.

Bias — an irrelevant first value distorts judgement.

Learning depends on the limited capacity of working memory, which can be overloaded.

The difficulty of the material itself.

Wasted load caused by poor presentation — clutter, confusing layout, distractions.

The useful effort of building understanding and connecting ideas into long-term memory.

Working memory can only hold a few items; too much and it can't be processed.

Cut extraneous load, keep intrinsic load manageable, and support germane load.

No — you can only manage load; capacity itself is fixed.

It directly improves teaching, instructions and design, backed by memory research.

The three load types are hard to measure separately and optimal load varies by person.

Measurement — the theory rests on the measurable limits of working memory.