Development of Mandarin Speech Test Materials for Civilian Pilots in China

Every time a civilian plane takes off, the only way pilots and air traffic controllers communicate—through air-ground radiotelephony—is a lifeline for flight safety. Without visual cues in the sky, clear hearing isn’t just a “nice-to-have”—it’s non-negotiable. But civilian pilots face constant noise exposure in cockpits, putting them at risk of hearing loss that could endanger everyone on board.

In China, the Civil Aviation Administration (CAAC) uses pure-tone audiometry (PTA) to check pilots’ “auditory fitness for duty.” PTA measures sensitivity to single sound frequencies (like a beep), but it has a critical flaw: it can’t predict whether a pilot can understand real radiotelephony speech—especially in noise. That’s why a team of researchers set out to create a Mandarin speech test tailored to the real-world communication needs of Chinese civilian pilots.

Why This Test Is Different: Real Radiotelephony, Real Tasks

The team’s goal was simple: build a test that reflects the actual language pilots use. They started with four trusted radiotelephony resources:

ICAO Radiotelephony Communication (the global standard for pilot-controller speech)
A Chinese radiotelephony training course
A guide for China’s pilot English proficiency exam
Emergency communication protocols

Working with captains and linguists, they narrowed down 12 “hearing-critical” task categories—from pre-flight checks and altitude changes to emergency situations. Examples include:

Departure: “Pushback approved, runway 36L.”
Flight Altitude: “Climb to 10,000 feet.”
Non-Routine Conditions: “Engine failure, requesting emergency landing.”

Next, they built 20 sentence lists following strict rules to ensure fairness and relevance:

All sentences use real radiotelephony language (no made-up phrases).
Tasks are balanced across lists (e.g., every list includes departure, altitude, and emergency phrases).
Sentences vary in length (3–13 Chinese characters) to avoid memory bias.
Keywords (the most important words for understanding) are simple (monosyllabic or two-syllable “spondees”) where possible.
No duplicate sentences.

Each list has 20 sentences and 100 keywords. To ensure quality, they recorded the sentences in an anechoic chamber (a room that eliminates echo) at the Chinese Academy of Social Sciences. The speaker? A 47-year-old male broadcaster with 20+ years of experience—his voice is clear, consistent, and familiar to pilots. After digital processing (to equalize loudness), the audio files were ready for testing.

Testing the Materials: Who We Studied and How

The team recruited 40 male student pilots from Shenzhen Airlines—all holding CAAC Class I certificates (the highest standard for civilian pilots). Here’s what you need to know about them:

Average age: 23.7 years (range: 21–26)
Mean flight time: 229.6 hours (range: 205–279)
Language skills: Fluent in English and Mandarin
Exclusions: No hearing loss in both ears, no history of ear disorders

Tests were conducted in a double-walled acoustic cabin that meets American National Standards Institute (ANSI) 2004 rules for audiometric rooms. The clinical audiometer (used to measure sound volume) was calibrated to International Electrotechnical Commission (IEC) 645-2:1993 standards—so results are reliable.

Pilots listened to sentences at six intensity levels (5–15 decibels hearing level, or dB HL) in 2dB steps. For each level, they repeated the keywords (e.g., “runway 36L” from a departure sentence). Their score? The percentage of correct keywords (word recognition score, WRS).

Key Findings: Reliable, Valid, and Tailored to Pilots

Here’s what the data revealed:

PTA vs. Speech Threshold: The pilots’ average PTA (hearing sensitivity) was 10.6 dB HL (mildly better than normal). The “speech reception threshold” (SRT)—the lowest volume where they understood 50% of keywords—was 8.22 dB HL. That’s consistent with PTA, meaning the test aligns with basic hearing ability.
WRS Variability: As volume increased from 5 to 15 dB HL, WRS jumped from 19.17% to 93.25%. But at the 50% threshold (8.22 dB HL), scores varied widely (±20%)—even though all pilots had normal hearing. Why? Familiarity with Mandarin radiotelephony: some pilots trained abroad had less practice with Chinese phrases.
Reliability & Validity: The 20 lists were highly consistent (Cronbach’s alpha = 0.981—where >0.8 is “good”). The remaining 14 lists (after removing those with non-monotonic scores, where WRS didn’t always improve with volume) had strong validity (Kaiser-Meyer-Olkin = 0.905, Bartlett’s test p < 0.001)—meaning they measure exactly what they’re supposed to: pilot speech recognition.

Compared to other Mandarin speech tests, this one is harder: the SRT is higher (8.22 vs. 3.1 or 6.3 dB HL) and the “slope” (how quickly scores improve with volume) is lower. That’s likely because our pilots had slightly higher PTA (mild hearing loss) and less familiarity with Mandarin radiotelephony.

What This Means for Pilot Safety

The big takeaway? This test is functional. While PTA checks “can you hear a beep?,” this test checks “can you understand a controller’s instruction in a quiet cockpit?” It’s based on the real language pilots use every day—not abstract sounds.

But there’s more work to do. The study tested pilots in quiet rooms—next, researchers need to run tests in real-world noise (like cockpit noise) to set pass/fail criteria for CAAC. Another key insight: flight experience matters. Even if a pilot has mild hearing loss, familiarity with radiotelephony might compensate—so we shouldn’t automatically ground experienced pilots with minor hearing issues.

The Bottom Line

This test is a critical step toward making CAAC’s auditory fitness checks more relevant. It’s not just about “hearing”—it’s about understanding. For pilots, that’s the difference between a smooth flight and a safety risk.

The study was published in the Chinese Medical Journal in 2019 by Mo-Sheng Hu (Beijing Tongren Hospital/Civil Aviation Medicine Center), Jing Chen (Beijing Institute of Otolaryngology), and colleagues. Funding came from the Safety Capacity Building Project of Civil Aviation (No. FSDSA0038).

doi:10.1097/CM9.0000000000000491

Was this helpful?

0 / 0