Sensation, Thresholds, and Signal Detection

6 min read Updated Apr 19, 2026

You are trying to hear a phone vibrate across the room while your roommate runs the vacuum. Sometimes you catch it. Sometimes you swear you heard it and check your phone to find nothing there. Sometimes you miss a real call. This is not a trivial, everyday annoyance. This is the textbook setup for everything in signal detection theory, and the MCAT loves it.

Before we get to detection, we need to define what “sensing something” even means.

Sensation vs. Perception

Sensation is the physical conversion of a stimulus into a neural signal. A rod in your retina flips a molecule, fires a neuron. That is sensation.

Perception is what your brain makes of that signal. Your brain decides the neural blip means “my phone is vibrating.” That is perception.

The physical step of converting stimulus energy into neural energy is called transduction. Rods and cones transduce light. Hair cells transduce sound. Free nerve endings transduce pain. Every sense has its own transducers.

Absolute Threshold of Sensation

Your phone vibrates so gently you catch it only half the time. That crossover point - the intensity at which you detect a stimulus 50% of the time - is your absolute threshold of sensation. Below the threshold, you would correctly detect the stimulus less than half the time (it is below the line you called “detectable”). Above the threshold, you would detect it more than half the time.

Stimuli below absolute threshold are called subliminal. They still reach your sensory organs, but you do not consciously detect them more than half the time.

Absolute threshold is not a hard, fixed number. It slides around based on:

Expectations. If you are waiting for a text, your threshold drops.
Experience. A new parent can hear a baby’s whimper through two closed doors.
Motivation. If you’re waiting to hear from someone important, you’re more sensitive to notifications.
Alertness. A drowsy person misses signals a well-rested person would catch.

Just Noticeable Difference and Weber’s Law

Absolute threshold asks, “Can you hear the phone at all?” The just-noticeable difference (JND) asks, “Can you hear the phone getting louder?” JND is the smallest change in a stimulus that a person can detect 50% of the time.

Here is the trick that the MCAT wants you to internalize: the JND is not a fixed amount, it is a fixed proportion. A 2 lb weight next to a 2.2 lb weight feels different - a 10% change. A 50 lb weight next to a 50.2 lb weight feels identical, because 0.2 is a tiny fraction of 50. To feel a change against 50 lb, you would need to add about 5 lb.

That proportionality is Weber’s Law:

$\frac{\Delta I}{I} = k$

where ΔI is the JND, I is the starting intensity, and k is a constant (the “Weber fraction”) specific to the sense. For weight, k is roughly 0.02 (a 2% change). For brightness, k is closer to 0.08.

Two horizontal rows of yellow circles. The top row grows by equal absolute increments (10, 20, 30, 40, 50, 60, 70 units) — each step looks smaller than the last. The bottom row grows by equal proportional increments (10, 14, 20, 27, 38, 54, 75 — each ~40% bigger than the previous) — each step looks like the same perceived increase — Weber's law in pictures. Top row: equal absolute increments produce shrinking perceived steps (the gap between 60 and 70 looks tiny). Bottom row: equal *proportional* increments feel like equal perceptual steps, because the JND scales with the baseline intensity. *Credit: MrPomidor via Wikimedia Commons (CC BY-SA 4.0).*

Signal Detection Theory

Now to your vacuum-vs-vibrating-phone problem. Signal detection theory (SDT) models how we decide whether a stimulus is present when conditions are noisy. Your brain is constantly trying to distinguish signal (the phone) from noise (the vacuum, the dog, the wind).

There are four possible outcomes on any trial:

	Signal Present	Signal Absent
You say “yes”	Hit	False Alarm
You say “no”	Miss	Correct Rejection

Hit. You say yes, the phone really was buzzing.
Miss. You say no, but the phone actually was buzzing. You missed the call.
False alarm. You say yes, but the phone never buzzed. You check and feel silly.
Correct rejection. You say no, and nothing was there.

d’ (Sensitivity) and c (Bias)

Signal detection splits performance into two independent pieces:

d’ (d-prime) measures sensitivity. How far apart are the “signal” and “noise” distributions in the brain? A bigger d’ means the signal stands out clearly from noise. A small d’ means they overlap and you cannot reliably tell them apart no matter how carefully you try.

c measures response bias (strategy). At what intensity do you decide to say “yes”? This is a personality trait of a detector, not a feature of the signal.

Conservative strategy. “I’ll only say yes when I’m 100% sure.” Many misses, few false alarms.
Liberal strategy. “I’ll say yes if I even suspect it.” Many hits, but also many false alarms.

A receiver operating characteristic (ROC) curve plot with false-positive rate on the x-axis (0 to 1) and true-positive rate on the y-axis (0 to 1). A blue dot at the top-left corner marks a perfect classifier. A red dashed diagonal line marks a random classifier. Three colored curves bow above the diagonal at increasing distances — labeled with arrows pointing 'Better' away from the diagonal and 'Worse' toward it — The ROC curve plots hit rate (true positives) against false-alarm rate (false positives) as the response criterion shifts from very conservative (bottom-left) to very liberal (top-right). The further the curve bows toward the upper-left corner, the higher the underlying sensitivity (d'). The diagonal is chance performance. *Credit: cmglee & MartinThoma via Wikimedia Commons (CC BY-SA 4.0).*

What is the absolute threshold of sensation?

Click to reveal answer

The minimum stimulus intensity needed to detect a stimulus 50% of the time. Below threshold = subliminal.

If you can just barely feel a 1 oz change on a 10 oz weight, how many ounces must you add to a 100 oz weight to feel a change? (Weber's Law)

Click to reveal answer

10 oz. Weber's Law says ΔI/I = constant.

\frac{1}{10}

= 0.10, so ΔI on a 100 oz weight is 0.10 × 100 = 10 oz.

In signal detection theory, what are the four outcomes?

Click to reveal answer

Hit (signal present, said yes), miss (signal present, said no), false alarm (signal absent, said yes), correct rejection (signal absent, said no).

What does a liberal response strategy in signal detection look like in terms of hits and false alarms?

Click to reveal answer

High hit rate AND high false alarm rate. A liberal subject says "yes" readily, catching real signals but also imagining phantom ones.