SLPOtherInternational journal of language & communication disorders2025

Establishing Norm of Connected Speech Measures for Descriptive Discourses in Cantonese-Speaking Adults.

Anthony Pak-Hin Kong, Chester Yee-Nok Cheung, Cherie Wan-Yin Wong

PMID 40384163

WHAT IT FOUND

Cantonese adults show varied speech patterns by age and education, not a simple decline.

Older adults struggled most with sequential picture tasks, while middle-aged adults peaked in procedural descriptions. High-education speakers produced more content but appeared less lexically diverse due to longer samples.

Key findings

01Older adults consistently showed poorer ability to convey information, particularly in sequential picture description tasks, while middle-aged adults often outperformed both younger and older groups in productivity.

02High-education speakers produced significantly more utterances and content units, but scored lower on type-token ratio and some efficiency measures because their longer samples diluted these metrics.

03The open-to-closed-class ratio showed a complex age-education interaction: low-education young and old speakers scored higher than high-education peers, but this effect reversed in the middle-aged group.

STILL TO COME

How it was doneWhat they foundWhat it means for SLPs

Read the rest of this summary

You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.

Already have one?

What it does not show

The study only included neurotypical adults, so these norms have not yet been validated against aphasia or dementia patients. Only one task per genre was used, which may not capture the full range of variation within that genre. Test-retest reliability was not assessed, so it is unknown if these measures are stable over time for monitoring progress. The sample was limited to Cantonese speakers in Hong Kong, so education cutoffs and linguistic norms may not apply to other Cantonese-speaking regions.

Declared interests

The authors declared no conflicts of interest. The data came from the Cantonese AphasiaBank, which is publicly accessible.

The easy way to misread this

Do not interpret lower type-token ratios or information units per word in highly educated patients as a sign of impairment. These scores are often lower simply because high-education speakers produce longer, more detailed samples, which dilutes these relative measures. Always look at absolute counts and consider the patient's education level before flagging a deficit.

Read it on PubMed →