Skip to content
PaperFren

Does hearing speech in noise predict story understanding in DLD?

Open paper intelligence

Among children with language disorder, how well they understood sentences in background noise predicted how well they understood spoken stories, but basic tone-hearing tasks did not.

Source

Speech Perception in Noise Predicts Oral Narrative Comprehension in Children With Developmental Language Disorder

Magimairaj BM, Nagaraj NK, Champlin CA, et al. · Frontiers in psychology · 2021

doi.org/10.3389/fpsyg.2021.735026Read the full paper ↗7 citationscc by

Study at a glance

Design
Cross-sectional — Single-time-point hierarchical regression within a DLD sample: age, non-verbal IQ and speech-in-quiet, then phonological STM, vocabulary, tone-perception thresholds and finally speech-in-noise predicting narrative comprehension
N
N=216 · 216 children with DLD recruited from nine school districts in Kansas and Texas; no typically developing comparison group
Population
School-age children (mean age about 90 months) meeting criteria for developmental language disorder with normal-range non-verbal IQ and hearing
Outcome
Comprehension of spoken stories (Test of Narrative Language comprehension questions)

Structured fields used in claim comparison tables when every cited study has a complete layer.

What they did

The researchers tested 216 school children with developmental language disorder on spoken story comprehension, nonword repetition (phonological short-term memory), two vocabulary tests, three tone-perception tasks (simultaneous masking, backward masking and frequency-sweep discrimination) and a sentence-in-noise test. They then used stepwise regression to see which abilities explained story comprehension after accounting for age, non-verbal IQ and sentence repetition in quiet.

What they found

Age, non-verbal IQ and speech-in-quiet together explained 36% of the variance in comprehension; phonological memory added 8% and vocabulary another 6%. The tone-perception tasks added nothing, but speech-in-noise ability explained a further 6% even after all other predictors, and this held when the order of the auditory measures was reversed. Tone-perception tasks were linked to non-verbal IQ, whereas speech-in-noise was not, suggesting tone tasks tap general attention while listening in noise draws on language-specific skills.

The limits

What it doesn't show

The design is correlational and at one time point, so it cannot show that poor listening in noise causes comprehension problems; both could reflect weaker language representations. There was no typically developing comparison group, so we cannot tell whether the relationship is special to DLD. The unique contribution of speech-in-noise was modest and its clinical significance was not quantified, stepwise regression results depend on entry order, and one author receives royalties from the comprehension test used.

Key terms

Developmental language disorder (DLD)
Persistent difficulty learning and using language that is not explained by hearing loss, low non-verbal intelligence or another known condition.
Speech perception in noise (SPiN)
The ability to understand speech against background noise, measured here as the quietest sentence level a child can repeat correctly.
Temporal processing deficit hypothesis
The proposal that language disorders arise from difficulty processing brief or rapidly changing sounds.
Backward masking
A task in which a brief tone is followed immediately by noise, making the tone harder to detect; used to probe auditory temporal processing.
Hierarchical regression
A regression in which predictors are entered in steps to see how much extra variance each set explains beyond those already entered.
Phonological short-term memory
Temporary storage of speech sounds, often measured by asking children to repeat made-up words.

Flashcards

1 / 10

0 of 10 answers reviewed

Research intelligence for this paper

See its role on concept claims, tensions it is part of, placement history, and related discoveries.

Open paper intelligence

Quiz yourself

1 / 6

Which measure uniquely predicted story comprehension after all other predictors?

Common questions

Isn't a sentence-in-noise test just another language test?

The authors address this by entering sentence repetition in quiet and vocabulary first, so speech-in-noise had to predict comprehension beyond language ability. The test sentences were also short and simple, designed for young children.

What does this mean for the idea that DLD comes from poor auditory timing?

Tone-based timing and frequency tasks did not add to predicting comprehension, which fits other large studies and failed training trials that challenge a purely auditory temporal-processing explanation.

What practical steps do the authors suggest?

Reducing classroom noise, using assistive listening devices such as remote microphones, and strengthening vocabulary and word retrieval, since listening in noise seems to matter for these children.

More on Language