The Ling 6 Sound Test

01What is this guide for?

If you work with a child who wears a hearing aid or a cochlear implant, one question has to be answered at the start of every session: is this child hearing today? The device may be in place, the battery may be full, the indicator light may be on — none of this proves that the child has access to speech sounds. A session run without verifying that the device works is a session run without knowing what the child hears that day.

The Ling 6 sound test answers this question in a minute or two and requires no equipment. Its simplicity is deceptive: applied correctly it screens access across the whole speech spectrum, applied carelessly it produces a reassuring but meaningless result. This guide covers what the test measures, how it is administered and how the results are interpreted.

02What is the Ling 6 sound test?

The test described by Daniel Ling (1926–2003) in Speech and the Hearing-Impaired Child (1976) originally consisted of five sounds; the addition of /m/ produced the six-sound version used today.

The idea behind it is straightforward: speech sounds are spread across a wide frequency range, and you do not have to test every phoneme to find out whether a person has access to that whole range. If you choose a small number of sounds that represent the low, mid and high frequency regions, you can screen quickly.

The six sounds are:

/m/ ("mmm") · /u/ ("oo") · /a/ ("ah") · /i/ ("ee") · /ʃ/ ("sh") · /s/ ("ss")

Three are vowels and three are consonants. The vowels are the corner vowels of the language; they mark the outer limits of the vowel space. The consonants represent three separate frequency regions from low to high.

The clinical value of the test comes not from a single measurement but from repetition. The aim is not to establish hearing thresholds but to establish the child's own usual response pattern and to catch departures from it early. This is why building the test into a daily routine is worth more than administering it occasionally and in detail.

03Which part of the audiogram does each sound test?

You cannot interpret a result without knowing which frequency region a sound probes. Rather than memorising individual numbers, it is enough to understand which part of the audiogram each sound corresponds to.

Formants in vowels

A vowel does not consist of a single frequency; it is made up of several simultaneous components. The two that matter clinically are the first formant (F1) and the second formant (F2). F1 is the stronger; F2 is roughly 15 dB quieter.

The two carry meaning together: a vowel can only be identified if both F1 and F2 are heard. When F2 is not heard the vowel does not disappear — it is heard as a different vowel.

The one exception is /u/, which can be identified from F1 alone. A correct response to /u/ therefore does not demonstrate that the child has access to the F2 region.

Why does F2 matter so much? For two reasons: it is markedly quieter than F1, so it is harder for a device to pick up and amplify; and most device users have poorer hearing at high frequencies than at low ones. Together these make the F2 region the most fragile part of access.

The speech banana

The banana-shaped area on the audiogram where the energy of speech sounds is concentrated is called the speech banana. Ling's six sounds mark this area from its low end to its high end. That is the visual logic of the test: if all six sounds are answered correctly, access across the whole speech banana is likely.

04What does the test measure, and what does it not?

Ling 6 can be applied at the first three levels of auditory perception:

Detection. "Is there a sound?" The child responds with a movement taught in advance.

Discrimination. "Are these two sounds the same or different?"

Identification. "What did you hear?" The child repeats the sound or points to its card.

It does not measure comprehension. Ling 6 is not a language or comprehension test; other tools are needed for that (Smiley, 2004).

Knowing the limits of the test matters as much as reading the result correctly:

It is not a diagnostic test. Ling 6 does not give hearing thresholds and does not replace audiological assessment. What it tells you is this: "something that was usual yesterday appears to have changed today." That is a signal to investigate, not a conclusion.

It does not cover all phonemes. Six sounds do not represent every phoneme in speech. The test is a quick check; it does not substitute for a full speech perception assessment.

A complete response does not prove that everything is in order. Even when all six sounds are repeated correctly, the result was obtained in a quiet room at about one metre. It does not mean access is maintained in classroom noise, at a distance, or for word-final consonants. Nor does it show that the device is performing as programmed in the clinic.

In short: Ling 6 answers "is there a problem?" quickly. It does not answer "is everything optimal?"

05How is the test administered?

The reliability of the test depends on procedural detail far more than people expect. The steps below should be followed in order.

Verify the device first

The Ling test does not replace the device check; it follows it. Before starting, check the battery or charge, the cleanliness and physical condition of the device, and perform a listening check where possible. Otherwise, if a sound is missed, you cannot tell whether the problem lies with the child or with the device.

Prepare the room and your position

The room should be quiet. Stand beside or behind the child; if your mouth is visible the test is invalid, because the child may respond by lip-reading.

Increase the distance in steps

Begin close to the child (about 20 cm). Once all sounds are answered, move to about one metre (90 cm); this is the one-to-one conversational distance and a suitable distance for the device microphone. Later you may move to two and three metres; three metres approximates listening in a group setting.

Present the sounds at the right level

Produce the sounds at ordinary conversational intensity. Do not exaggerate. This applies to /s/ in particular: in natural speech /s/ is markedly quieter than the other sounds. If it is stressed, the sound becomes distorted, low-frequency components are introduced, and the test no longer says anything about high-frequency access.

Change the order and the intervals every time

This is the most frequently omitted detail and the one that most often invalidates the result.

• Present the sounds in a different order on each round.
• Use some sounds more than once; make nine or ten presentations per round, not six.
• Vary the interval between presentations.

The reason: if order and timing are fixed, the child notices that a sound has been missed and starts to guess. A child who does not hear /s/ at all may infer the gap from the rhythm and answer "ss". The test then shows access that is not there.

Give no cues

• Do not say "I am going to make a sound now." Failing to notice that a sound was presented is itself the evidence that access is absent in that region; warning the child destroys that information.
• Do not ask "did you hear that?" The question already tells the child that a sound was presented and makes the response unreliable. Part of the test is listening to how the sound is repeated.
• Do not give feedback during the test. Responses such as "well done, correct" leak cues and break concentration. Save reinforcement for the end.

Do not break the flow when a sound is missed

If the child does not respond to a sound, immediately present a sound from a different frequency region, so that the child does not realise a sound was missed. Return to the missed sound later in the round and see whether the response repeats.

Record the response exactly as given

Marking correct or incorrect is not enough. Write down exactly what the child said. If they said "oo" instead of "ee", record that; the whole of the interpretation rests on this detail. Note separately where no response was obtained.

Establish the baseline

In the first days the aim is not to look for problems but to establish the child's baseline: which sounds are answered, at what distance, and how consistently. Once this pattern is known, the test becomes a monitoring tool that reveals departures from it. Without a baseline there is no such thing as a "poor result".

06Interpreting the results

The least appreciated feature of Ling 6 is that errors are not random. Particular error patterns point to particular regions of the audiogram.

The classic error: "oo" for "ee"

If a child answers "oo" when /i/ is presented, this is not an articulation problem or inattention. The explanation is this: the F2 component of /i/ lies in the high-frequency region and was not heard. The remaining F1 component sits in almost the same region as the F1 of /u/. The child has reported what they heard correctly; what they heard really did resemble /u/.

This single observation indicates a problem with high-frequency access, and it is usually seen together with a missed /s/.

Error patterns and what they may mean

A correct response does not give a threshold

In pure-tone audiometry, responding to a stimulus (pressing a button, turning towards the sound) is possible at threshold. Recognising and repeating a speech sound requires more: the sound must be heard, discriminated and identified. A sound that is repeated correctly therefore does not show where hearing actually is; it shows the limit of "it can be no worse than this".

The practical consequence: what the Ling test yields is not a threshold but a lower bound. The child's hearing may be better than the result suggests; it cannot be worse.

What to do when the pattern persists

An error seen in a single session may reflect distraction or fatigue. What matters is whether the pattern repeats.

1. Repeat the device check and work through troubleshooting.
2. Repeat the test in the next session and see whether the same sound or sounds are missed again.
3. If the same error persists for two to three consecutive days, inform the family and refer to the audiologist.

A recurring error pattern may reflect a device fault or a programming problem, a middle-ear condition (including transient states such as otitis media with effusion), a permanent change in hearing level, or a need to update the cochlear implant map.

07Adapting to age and developmental level

The test is administered with different response formats depending on the child's age and level of auditory development.

With infants, observation is limited to the detection level: startle, stopping or resuming sucking, turning towards the sound, widening of the eyes. These responses are subjective, which is why regular observation by the same person is valuable.

With young children, an observable response is established through conditioning. The Turkish Ministry of National Education support programme describes this approach directly: the child is asked to give a behavioural response on hearing the sound — dropping a block into a basket, for example. Once conditioning is complete, the child begins to respond to sound spontaneously.

With children who cannot repeat, pictures or toys representing the sounds are used and the child points to the one they heard. This makes it possible to measure the identification level in children whose articulation has not yet developed.

With children who have articulation difficulties, a response does not have to be produced perfectly to count as correct. What matters is that the child reports which sound they heard consistently. If the tester cannot recognise the child's production reliably, the test can be used at the detection level only.

With children who have additional needs, pointing to pictures may add a cognitive load. In that case, working at the detection level in a wait–listen–respond format is enough to show that the child has access to the speech frequencies.

08Common administration errors

What these errors have in common is that they turn the test into something that looks better but tells you less.

09The current state of the test: from Ling 6 to LMH

Ling's six sounds have been in use for more than forty years. With the spread of digital hearing aids and cochlear implants, however, the question of whether six sounds provide enough information has been raised.

What was proposed. Jane Madell and Joan Hewitt reviewed more than 230 cochlear implant programming sessions retrospectively and reported that the most common speech perception errors were not captured by the six sounds. They therefore added four further consonants: /z/, /h/, /n/ and /dʒ/. The resulting LMH 10 Sound Quick Test takes its name from Ling-Madell-Hewitt and from the low, mid and high frequency regions it screens. The added sounds are intended mainly to increase mid-frequency information (Madell & Hewitt, 2022).

LMH is not a single test but a three-level battery: the 10-sound quick test, followed by imitation of all phonemes, and finally identification of the medial consonant in vowel-consonant-vowel sequences (aba, ata, afa).

What was criticised. The proposal has not been received without debate. The educational audiologist Peter Keen has challenged its rationale (Keen, 2022). His main points are:

• The criticism that the six sounds are memorised applies not to the test itself but to poor administration; memorisation is impossible when order and intervals change on every round.
• Ling 6 is a quick check by design; it never claimed to cover all phonemes. Speech perception problems belong to the programming process.
• If the aim is diagnostic, ten sounds are not sufficient either; a range appropriate to the child's developmental level is required.
• Among the chosen sounds, /h/ is very quiet in ordinary speech and unlikely to be picked up by a hearing aid; produced loudly, it distorts. /dʒ/ is a sound that typically developing children begin to use at around four years of age, which limits its applicability at younger ages.

Keen's recommendation is to keep Ling 6 as the daily quick check and to add weekly, broader speech discrimination assessments.

What this means for this guide. Both sides agree on one point: Ling 6 retains its place as a daily monitoring tool but does not replace a comprehensive speech perception assessment. Because common practice in Türkiye continues to be based on the six sounds, this guide takes the six sounds as its basis. The LMH battery is worth knowing as a current development to follow.

10The Ling sounds in Turkish

The Ling sounds are language-specific. The test was developed for North American English, and two questions have to be asked when it is transferred to another language: are these sounds present in that language, and are their acoustic properties the same?

Are the sounds present? For Turkish the answer is yes. All six Ling sounds occur in Turkish: m, u, a, i, ş, s. This is not true of every language; in Māori, for example, /s/ and /ʃ/ do not exist, so the test can only be used at the detection level and discrimination and identification remain invalid (McDonnell, 2014). The four sounds added by LMH (z, h, n, c) are also present in Turkish.

Are the acoustic properties the same? Here care is needed. Vowel formant values vary between languages; the acoustic properties of Turkish /a/, /i/ and /u/ do not correspond exactly to those in English. Turkish also has eight vowels, among them the front rounded vowels (ö, ü) that have no English counterpart. Ling's three vowels mark the corners of the vowel space in Turkish as well, so the logic of the test holds; but the Turkish vowel space is more crowded.

The phonetic and phonological properties of the Turkish Ling sounds have been examined in detail and compared with the Ling sounds of other languages by Önen and Kemaloğlu (2022). Clinicians administering the test in Turkish are advised to consult this source.

Accent and speaker differences. The tester's voice, sex and speech habits affect the acoustic properties of the sounds to some degree. Vowel components sit at lower frequencies in male voices and at higher frequencies in female and child voices. This variability does not invalidate the test; but having the test administered by the same person wherever possible improves the consistency of monitoring.

11Summary

Ling 6 checks that the child is hearing, not that the device is working. It follows the device check; it does not replace it.

The six sounds screen the speech spectrum from low to high. Rather than memorising values, it is enough to know which region of the audiogram each sound probes.

The test is applied at the detection, discrimination and identification levels; it does not measure comprehension and it is not a diagnostic test.

Procedural detail determines the result: order and intervals must change on every round, sounds must not be exaggerated, no cues may be given, and responses must be recorded exactly as produced.

Errors are not random. An "oo" response to /i/ indicates a problem with high-frequency access; comparable patterns point to particular regions of the audiogram.

The value lies in repetition. Without a baseline, departures cannot be recognised. A recurring error pattern that persists for two to three days calls for referral to the audiologist.

All six sounds exist in Turkish, but their acoustic properties vary by language; for a Turkish-specific account, Önen and Kemaloğlu (2022) should be consulted.

12References

Keen, P. (2022). The Ling 6 Sound Test: A day-to-day hearing aid check or a frequency-specific analysis of hearing aid provision. BATOD Magazine, May 2022, 40–55.

Ling, D. (1976). Speech and the hearing-impaired child: Theory and practice. Washington DC: A. G. Bell Association for the Deaf.

Ling, D. (1989). Foundations of spoken language for hearing-impaired children. Washington DC: A. G. Bell Association for the Deaf.

Madell, J., & Hewitt, J. (2022). The Ling-Madell-Hewitt (LMH) Test Battery. Volta Voices.

McDonnell, S. (2014). The Ling Sound Test: What is its relevance in the New Zealand classroom? Kairaranga, 15(2), 48–55.

Ministry of National Education. (2021). Support education programme for individuals with hearing impairment. Ankara: Directorate General for Private Education Institutions.

Önen, Ç., & Kemaloğlu, Y. K. (2022). Phonetic and phonological analysis of the Turkish Ling sounds. In Y. K. Kemaloğlu, H. Kasapoğlu Çengel & G. Mengü (Eds.), The importance of phonetic and phonological features of speech in audiology: The example of the Turkish language (pp. 117–121). Ankara: Türkiye Klinikleri.

Scollie, S., & Glista, D. (2012). Stimuli and normative data for detection of Ling-6 sounds in hearing level. American Journal of Audiology, 21(2), 232–241.

Smiley, D. F. (2004). Using the Ling 6-sound test everyday. AudiologyOnline.