Sensation & Perception

Sensation & Perception

Part 1 of 4 — Thresholds, Weber's Law & Signal Detection

Sensation vs. Perception (the core distinction)

• Sensation: transduction of physical stimuli into neural signals by sensory receptors — a bottom-up, data-driven process
• Perception: the brain's organization and interpretation of those signals — shaped by top-down expectations, context, and experience
• One physical stimulus can yield different percepts (ambiguous figures); different stimuli can yield the same percept (constancies)

Psychophysics: Quantifying the Stimulus–Experience Link

Psychophysics — founded by Gustav Fechner (building on Ernst Weber) — measures how physical intensity maps onto subjective experience.

• Stevens' Power Law: $S = k I^n$ — handles modalities Fechner's law fits poorly (e.g., the sensation of electric shock grows faster than intensity, $n > 1$).

Why Weber's Law matters: a 1-lb change is obvious when lifting 5 lbs but invisible when lifting 100 lbs. Detectability depends on the ratio, not the absolute difference.

Signal Detection Theory (SDT)

Classical thresholds treat detection as purely sensory. SDT recognizes that detection is also a decision made under uncertainty, influenced by motivation and expectations. It separates:

• Sensitivity ($d'$) — how well the observer discriminates signal from noise
• Response criterion (bias, β) — how willing the observer is to say "yes"

• Liberal criterion → more hits and more false alarms
• Conservative criterion → fewer false alarms but more misses
• Payoffs, fatigue, and expectations shift the criterion without changing $d'$ — this is why raw hit rate alone never proves better perception

Sensory Adaptation (a receptor-level change)

• Sensory adaptation: receptors respond less to a constant, unchanging stimulus over time (you stop feeling your watch). It is a peripheral, receptor-level change.
• Distinguish it from habituation, a learned, central decrease in response to a repeated stimulus — habituation shows dishabituation (response recovery when a novel stimulus intervenes), adaptation does not depend on learning.
• Pain adapts poorly — a protective design feature.

Transduction: the Common First Step

Every sense converts a physical stimulus into a receptor potential, then into action potentials relayed (for all senses except smell) through the thalamus to a dedicated cortical area. Later parts trace each pathway.

Thresholds & Signal Detection 🎯

Worked Examples — Thresholds & SDT

Question: A subject just detects a brightness change from 120 cd/m² to 126 cd/m². What change is needed at 300 cd/m²?

Solution: $k = \frac{126 - 120}{120} = 0.05$ $\Delta I = 0.05 \times 300 = 15 \text{ cd/m}^2 \checkmark$

Key idea: Compute the Weber fraction from the given pair, then scale. The answer is never "the same 6 units."

Question: Observer A: 85 hits / 40 false alarms per 100 trials of each type. Observer B: 70 hits / 10 false alarms. Who is more sensitive? Who is more liberal?

Solution:

1. Sensitivity tracks the gap between hit and false-alarm rates. A: $0.85 - 0.40 = 0.45$; B: $0.70 - 0.10 = 0.60$ → B discriminates better.
2. A says "yes" far more often overall → A holds the more liberal criterion. ✓

MCAT lesson: A high hit rate with a high false-alarm rate reveals bias, not skill.

Question: A word flashed too briefly for participants to report still speeds recognition of related words. Does this show subliminal stimuli control behavior?

Solution:

1. The flash is below the absolute threshold for conscious report, yet produces priming — a real but small, short-lived effect on processing speed. ✓
2. It does not demonstrate durable attitude or behavior change; controlled studies find no strong "subliminal persuasion."

Connection: Priming is top-down facilitation; the MCAT rewards the modest, evidence-bounded conclusion.

Key Takeaways — Part 1

• Sensation = bottom-up transduction; perception = top-down interpretation.
• Absolute threshold and JND are both defined at 50% detection; Weber's Law: $\Delta I / I = k$.
• SDT splits detection into sensitivity ($d'$) and criterion (bias); payoffs move the criterion, not $d'$.
• Sensory adaptation is receptor-level and stimulus-bound; habituation is learned and shows dishabituation.

• Low-yield extras: Stevens' power law ($S = k I^n$, with $n > 1$ for electric shock) as the alternative to Fechner's logarithmic law

Sensation & Perception

Part 2 of 4 — Vision & Hearing: From Transduction to Cortex

Vision: Structure of the Eye

Light path: cornea (most refraction) → pupil (iris controls size) → lens (accommodation: ciliary muscles change lens shape to focus) → retina.

• Approximate counts: about 120 million rods vs. about 6 million cones per retina.

Phototransduction (know the direction of the change)

1. Light isomerizes 11-cis-retinal → all-trans-retinal, activating the opsin.
2. Activated pigment stimulates transducin (a G protein) → phosphodiesterase → cGMP falls.
3. cGMP-gated Na⁺ channels close → the photoreceptor hyperpolarizes → less glutamate released.

Trap: photoreceptors are depolarized in the dark ("dark current") and hyperpolarize in the light — the reverse of a typical receptor.

Visual Pathway & Parallel Processing

Photoreceptors → bipolar cells → ganglion cells (axons form the optic nerve) → optic chiasm (nasal fibers cross, so each visual field maps to the opposite hemisphere) → lateral geniculate nucleus (LGN) of the thalamus → primary visual cortex (V1), occipital lobe.

• Parvocellular pathway: fine spatial detail and color; best with stationary objects
• Magnocellular pathway: motion and coarse, fast processing
• Feature detectors (Hubel & Wiesel): V1 neurons tuned to edges and orientations

• Hubel and Wiesel shared the 1981 Nobel Prize for this work.

Two Correct Theories of Color

• Trichromatic theory (Young–Helmholtz): three cone types (S/M/L wavelengths) — true at the receptor level
• Opponent-process theory (Hering): red–green, blue–yellow, black–white paired channels in ganglion/LGN cells — explains afterimages (fatigue one side, rebound to the other)
• MCAT answer: both are right, at different stages of processing

Hearing: Structure and Transduction

Sound path: pinna → auditory canal → tympanic membrane → ossicles (malleus → incus → stapes, amplifying pressure) → oval window → cochlear fluid → basilar membrane → hair cells (in the organ of Corti) bend against the tectorial membrane → K⁺ influx depolarizes the hair cell → neurotransmitter onto auditory nerve (CN VIII) → brainstem → medial geniculate nucleus (MGN) of the thalamus → primary auditory cortex (A1), temporal lobe.

Trap: hair-cell depolarization is by K⁺ influx (the endolymph is unusually K⁺-rich) — an exception to the "Na⁺ depolarizes" rule.

Pitch Coding

Balance: the Vestibular System

• Semicircular canals (three, orthogonal): rotational acceleration
• Utricle & saccule (otolithic organs): linear acceleration and head tilt, via calcium-carbonate otoliths bending hair cells

Vision & Hearing 🎯

Worked Examples — Vision & Hearing

Question: A stroke patient loses vision in the entire left visual field of both eyes. Where is the lesion?

Solution:

1. At the optic chiasm, nasal retinal fibers cross, so all information from the left visual field is routed to the right hemisphere. ✓
2. Loss of one full visual field (homonymous hemianopia) → lesion behind the chiasm on the right (right optic tract, LGN, or V1).

Key idea: Eyes ≠ visual fields. Chiasm lesions cut peripheral (temporal) fields of both eyes; post-chiasm lesions cut one whole field.

Question: A retinal disease destroys cones but spares rods. Predict the patient's vision.

Solution:

1. No cones → no color vision and poor acuity, especially at the fovea (central blindness in daylight). ✓
2. Rods intact → preserved dim-light and peripheral vision, though everything is grayscale and blurry.

Connection: The reverse pattern (rod loss) → night blindness and tunnel vision with preserved central color vision.

Question: Patient X has fused ossicles; patient Y has destroyed cochlear hair cells. Contrast their deficits and the step of the pathway lost.

Solution:

1. X: sound cannot be mechanically amplified to the oval window → conductive loss; transduction machinery is intact, so bone conduction still works. ✓
2. Y: transduction itself fails — no receptor potential regardless of amplification → sensorineural loss; hair cells do not regenerate.

MCAT lesson: Identify which step (conduction vs. transduction vs. neural relay) an experimental manipulation removes.

Key Takeaways — Part 2

• Light hyperpolarizes photoreceptors (cGMP falls, channels close); dark = depolarized.
• Pathways: retina → LGN (thalamus) → V1; cochlea → MGN → A1. Smell is the only thalamus-bypassing sense.
• Color: trichromatic at the cones, opponent-process afterward (explains afterimages).
• Pitch: place theory for high frequencies, temporal/frequency coding (+ volley) for low.
• Vestibular: semicircular canals = rotation; utricle/saccule = linear acceleration and tilt.
• Rods far outnumber cones and dominate the periphery; cones cluster in the fovea.

• Low-yield extras: about 120 million rods vs. 6 million cones per retina; Hubel and Wiesel's 1981 Nobel Prize

Sensation & Perception

Part 3 of 4 — Body Senses, Gestalt Organization & Depth

Somatosensation: the Skin Senses

• Signals ascend to the somatosensory cortex (parietal lobe), mapped as a distorted homunculus — cortical area tracks receptor density (fingers, lips), not body-part size.
• Two-point threshold: minimum separation felt as two touches; smallest where receptor density is highest.

Gate Control Theory of Pain (Melzack & Wall)

• A spinal-cord "gate" modulates whether nociceptive signals reach the brain.
• Activity in large-diameter touch fibers closes the gate on small-diameter pain fibers — why rubbing a stubbed toe helps.

• Fiber classes: the large touch fibers are A-beta; the small pain fibers are A-delta (fast, sharp pain) and C (slow, dull or burning pain). Melzack and Wall proposed the theory in 1965.

• Descending signals from the brain (attention, expectation, endorphins) also gate pain — the mechanism behind placebo analgesia and athletes not noticing injuries mid-game.

Smell & Taste (the chemical senses)

• Olfaction: odorants bind olfactory receptor neurons → olfactory bulb → piriform cortex/limbic system, bypassing the thalamus (unique) — one reason odors evoke vivid emotional memories.
• Gustation: taste buds (housed in the tongue's papillae) detect five basic tastes — sweet, sour, salty, bitter, umami; signals travel via cranial nerves → brainstem → thalamus → gustatory cortex (insula).

• The taste nerves: CN VII (anterior two-thirds of the tongue), CN IX (posterior third), CN X (epiglottis).

• Kinesthesia/proprioception: muscle-spindle and joint receptors report limb position — distinct from vestibular balance.

Gestalt Principles ("the whole is other than the sum of its parts")

The umbrella law is Prägnanz — we perceive the simplest organization.

• The Gestalt school was founded by Max Wertheimer, Wolfgang Köhler, and Kurt Koffka.

Depth Perception

• Binocular cues (need two eyes): retinal disparity (image difference between eyes; larger = closer) and convergence (eyes rotate inward for near objects)
• Monocular cues (one eye suffices): relative size, interposition (overlap), linear perspective, texture gradient, light/shadow, motion parallax (near objects sweep past faster), and accommodation
• Visual cliff (Gibson & Walk): crawling infants refuse the "deep" side → depth perception emerges early, by crawling age

Perceptual Constancies & Top-Down Effects

• Size, shape, and color constancy keep objects stable despite changing retinal images — misapplied constancy underlies size illusions (Ponzo, Müller-Lyer).
• Perceptual set: expectations prime interpretation (context effects, priming).
• Bottom-up = feature-driven; top-down = knowledge-driven; normal perception blends both.

Body Senses & Perceptual Organization 🎯

Worked Examples — Organization & Depth

Question: A researcher measures two-point thresholds on the fingertip and the calf. Predict the result and explain it in terms of receptors and cortex.

Solution:

1. Fingertip threshold ≪ calf threshold — two nearby points on the fingertip are felt as two, but must be far apart on the calf. ✓
2. Reason: fingertips pack a high density of small-receptive-field mechanoreceptors and command a disproportionately large slice of the somatosensory homunculus.

Key idea: Acuity tracks receptor density and cortical magnification, not body-part size.

Question: Viewers see a triangle in the "Kanizsa figure" even though only three Pac-Man-shaped wedges are drawn. Which principle, and what does it show about perception?

Solution:

1. Closure — the visual system fills the gaps to complete a simple whole (illusory contours). ✓
2. It shows perception is constructive: the percept contains structure not present in the stimulus (top-down completion consistent with Prägnanz).

Connection: Contrast with proximity/similarity, which group given elements rather than inventing contours.

Question: In the Ponzo illusion, two identical bars drawn over converging "railroad tracks" look unequal. Explain using constancy.

Solution:

1. Linear perspective signals that the upper bar is farther away. ✓
2. Size constancy scales perceived size by assumed distance: same retinal size + greater assumed distance → perceived as larger.

MCAT lesson: Illusions are not receptor failures; they are normally adaptive top-down rules applied to a tricky stimulus.

Key Takeaways — Part 3

• Match mechanoreceptors to stimuli (Pacinian = deep pressure/vibration; Meissner = light touch; Merkel = sustained texture; Ruffini = stretch; free endings = pain/temperature).
• Gate control: touch-fiber and descending activity close a spinal gate on pain.
• Smell bypasses the thalamus → strong emotion/memory links; taste does not.
• Binocular cues = retinal disparity + convergence; everything else (including motion parallax) is monocular.
• Constancies stabilize perception — and generate illusions when their assumptions are violated.

• Low-yield extras: gate-control fiber classes (A-beta touch; A-delta and C pain) and the 1965 date; the Gestalt founders Wertheimer, Köhler, and Koffka; the taste cranial nerves CN VII (anterior two-thirds), IX (posterior third), X (epiglottis)

Sensation & Perception — Deep Dive

Part 4 of 4 — MCAT Integration: Top-Down Perception, Streams & Reading Psychophysics Passages

Bottom-Up Meets Top-Down: The Full Circuit

Parts 1–3 built the pipeline: stimulus → receptor transduction → thalamic relay (except smell) → primary cortex → association cortex. Part 4 is about how the MCAT tests the interaction between that data-driven stream and the brain's expectations:

• Perceptual set: expectations prime one interpretation of an ambiguous input (the same figure reads as "13" in a number series and "B" in a letter series)
• Context effects: identical gray patches look different in different surrounds — the brain reports its inference about the object, not the raw receptor output (this is lightness constancy doing its job)
• Priming and motivation act on interpretation and on the response criterion — recall from Part 1 that expectations move β, not $d'$

The Two Cortical Visual Streams

A patient who can describe an object but cannot reach for it accurately (or vice versa) is a stream-dissociation item — match the deficit to the pathway, and remember the magnocellular input (motion, Part 2) feeds mainly dorsal while parvocellular (detail, color) feeds mainly ventral.

How Psychophysics Passages Present Data

1. Detection tables (hits/false alarms): always compute both rates. Hit rate alone confounds sensitivity with criterion.
2. Threshold-vs-intensity plots: a JND that grows linearly with baseline intensity is Weber's Law, $\Delta I / I = k$; a constant JND violates it.
3. Adaptation paradigms: prolonged exposure → reduced response → aftereffect in the opposite direction (motion aftereffect, color afterimages). Aftereffects are the standard evidence that a dedicated, fatigable channel exists for that feature.
4. Habituation designs in infants: looking time falls with repetition, then recovers to a stimulus the infant treats as novel — perception researchers use recovery as evidence the infant discriminates the two stimuli.

The Classic Discrimination Traps (final pass)

• Sensory adaptation (receptor-level, unlearned) vs. habituation (central, learned, shows dishabituation)
• Sensitivity ($d'$) vs. criterion (β) — payoffs and instructions move only the criterion
• Trichromatic (cone receptor level) vs. opponent-process (ganglion/LGN level; explains afterimages) — both are correct, at different stages
• Place coding (basilar-membrane location, high frequencies) vs. frequency/temporal coding (firing rate matches the tone, low frequencies; volley principle in between)
• Weber's Law is about proportional change; never answer with a constant absolute increment

Worked Example — An Expectation-and-Detection Passage, Fully Read

Passage. Radiology trainees view chest images, half containing a subtle nodule. In Block 1, trainees are told nodule prevalence is 10%; in Block 2, a different image set with the same objective prevalence (50% of trials) is described as coming from a "high-risk clinic." Results: Block 1 — 62% hits, 8% false alarms. Block 2 — 84% hits, 24% false alarms. Computed $d'$ does not differ between blocks. A final experiment finds trainees fixate the nodule region equally long in both blocks, but report "no nodule" more often in Block 1 even after fixating it.

Step 1 — classify the manipulation. Objective stimuli are matched; only the described prevalence changes. This manipulates expectation, a top-down variable — nothing about receptors or acuity can differ between blocks.

Step 2 — read the SDT signature. Hits AND false alarms rise together while $d'$ stays flat: a liberal criterion shift. The trainees did not become better discriminators; they became more willing to say "nodule." Any answer choice claiming improved sensitivity, lowered absolute threshold, or sharpened perception is wrong on the numbers given.

Step 3 — use the eye-tracking dissociation. Equal fixation with different reports shows the information reached the eyes in both blocks; the difference arose at the decision/interpretation stage. This is the passage's way of localizing the effect downstream of sensation — exactly the sensation-vs-perception distinction from Part 1, now measured.

Step 4 — anticipate the extension item. "Which change would most likely increase $d'$ rather than shift criterion?" Criterion movers: instructions, payoffs, prevalence beliefs. Sensitivity movers: better image contrast, longer viewing time, training that improves discrimination. Match the intervention to the SDT parameter it touches — that mapping is the single most reused P/S data skill.

Integration: Streams, SDT & Psychophysics Data 🎯

Key Takeaways — Part 4

• Perceptual set, context, and prevalence beliefs act top-down: they shift interpretation and response criterion, never $d'$ — hits and false alarms rising together is a criterion shift
• Ventral stream (occipital to temporal) = identity/"what" (agnosias); dorsal stream (occipital to parietal) = location, motion, action/"where-how"; magno feeds dorsal, parvo feeds ventral
• Data-reading toolkit: always pair hit rate with false-alarm rate; JND proportional to baseline = Weber's Law; aftereffects = fatigable dedicated channel; infant looking-time recovery = discrimination evidence
• Final trap list: adaptation vs habituation (dishabituation decides), trichromatic vs opponent (receptor vs ganglion stage), place vs frequency coding (high vs low pitch), sensitivity vs criterion (what the intervention touches)
• Passage habit: classify every manipulation as bottom-up (stimulus, receptors) or top-down (expectation, payoff, context) BEFORE reading the answer choices