Conductive hearing loss affects loudness perception; sensorineural loss also degrades frequency resolution

What it is

Hearing loss is not one thing. Clinically it splits into two fundamentally different pathological routes: conductive hearing loss (a problem in the outer or middle ear — earwax blockage, middle-ear fluid, otosclerosis) and sensorineural hearing loss (a problem in the cochlear hair cells or auditory nerve — most age-related and noise-induced hearing loss falls here). The two affect sound in completely different ways: conductive loss roughly acts like an attenuator bolted onto the ear, scaling every sound down proportionally; sensorineural loss doesn't just turn sound down — it also damages the ear's own ability to resolve fine frequency detail.

This distinction leads directly to a counterintuitive conclusion: raising volume does not compensate for sensorineural hearing loss. Turning the volume back up largely fixes conductive loss; with sensorineural loss, even sound played louder than a normal-hearing person would need can still be hard to understand, because the problem isn't "loud enough" — it's "clear enough to tell apart."

Why it happens

Conductive loss occurs somewhere along the path before sound reaches the inner ear; the cochlea and auditory nerve themselves are usually normal, and the sound is simply attenuated as a whole along that path. Once that attenuation is made up for (by amplification), the normal downstream cochlear analysis and neural transmission proceed as usual, and intelligibility is largely restored — which is why simple amplification is often effective for this type of loss.

Sensorineural loss occurs inside the cochlea itself, and when the outer hair cells — responsible for converting mechanical vibration into neural signals and performing an initial frequency analysis — are damaged, the cochlea's tuning across different frequencies becomes coarse: neighboring frequency components that should be processed separately now blur together. This drop in frequency resolution is especially damaging because the consonants that carry speech's highest semantic discriminability depend precisely on fine spectral detail (telling apart minimal pairs relies on exactly this kind of detail). Amplifying the sound only solves "can it be heard at all" — it does nothing for "it's all smeared together and can't be told apart." A damaged cochlea also often comes with recruitment: the volume range between "just barely audible" and "uncomfortably loud" is significantly compressed, so simple linear amplification tends to suddenly make a barely-audible sound feel harsh, while the usable, comfortable loudness range in between is narrow.

Studying it

Clinically, air conduction and bone conduction thresholds are tested separately to distinguish the two types: if the air conduction threshold is markedly worse than the bone conduction threshold (an air-bone gap), it points to a problem in the sound-conduction pathway — conductive; if air and bone conduction thresholds rise together by a similar amount, it points to a problem in the cochlea or nerve itself — sensorineural. Frequency resolution requires a separate psychoacoustic test (e.g., having a participant discriminate whether two closely spaced frequencies differ) — a different measure from a plain pure-tone threshold test.

Where it stops holding

• A large share of real-world hearing loss is mixed (both a conductive and a sensorineural component); not every case sorts cleanly into one category.
• The degree and type of hearing loss are often distributed unevenly across frequency; the same person may show different degrees, or even different types, of loss at different frequencies — an audiogram is typically a curve that varies with frequency, not a single number.
• Age-related sensorineural loss (presbycusis) typically starts at high frequencies and progressively worsens; this pattern is documented elsewhere and isn't repeated here.

Applying it

• Don't treat "turn up the volume" as the default accessibility fix for hearing: it only helps with conductive loss or the "can't hear it at all" component of loss, and does nothing for the common sensorineural pattern of "can hear it but can't make it out" — it can even cause discomfort due to recruitment.
• When evaluating the accessibility of interface audio content, test against a hearing-loss profile that simulates reduced frequency resolution (rather than a simple overall attenuation filter), which is what actually reflects a sensorineural-loss user's experience.
• When designing critical alert sounds or speech content, avoid making key distinguishing information depend only on subtle spectral differences (e.g., two alert tones that differ only in fine harmonic detail) — sensorineural-loss users may not be able to reliably tell such differences apart.

Related

• Same group: A3.17.2 Unilateral hearing loss completely eliminates localization ability that depends on interaural time and level differences · A3.17.3 Compression in hearing devices alters the intended loudness hierarchy of digital alert sounds · A3.17.4 A normal audiogram with difficulty understanding speech in noise is an easily overlooked hearing impairment
• Nearby: A3.08 Implications of hearing loss for interaction design · A3.01 Audible frequency range
• Search terms: conductive hearing loss · sensorineural hearing loss · recruitment · frequency resolution