SLPOtherAutism : the international journal of research and practice2021

Psychometric properties of the Chinese Parent Version of the Autism Spectrum Rating Scale: Rasch analysis.

Weili Yan, Richard J Siegert, Hao Zhou and 21 others

PMID 33845648

WHAT IT FOUND

A Chinese parent-rated autism scale can be shortened into a total score and three subscales that work consistently and show no scoring differences tied to age, sex, diagnosis status, rater or city.

The converted score did not diagnose better than the raw score.

Key findings

01The original 71-item Chinese parent scale did not fit the Rasch model well, and 22 of its items showed poor individual fit.

02After combining similar items and removing problematic items, the social communication, self-regulation, unusual behaviour and total scales showed good fit in a 250-child subsample and no bias remained for age, sex, case status, rater or city.

03The transformed score did not improve diagnostic accuracy: raw scores had an area under the curve of 0.835, and transformed scores had 0.834.

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

Only children aged 6 to 13 were studied, so it does not tell therapists how the scale behaves for adolescents aged 14 to 18 or younger children aged 2 to 5. The study did not collect language level, IQ or adaptive functioning, so it cannot show how the scale relates to those clinical features. Only the parent version was analysed, so teacher ratings were not tested. Good fit required combining similar items and deleting items, so the final scales are not the original 71-item scale used as a simple total score. The diagnostic comparison was cross-sectional, and the authors say the advantage of transformed scores may be clearer only with repeated measures over time. There were some floor and ceiling effects, with a small ceiling effect on social communication and some floor effect on all three subscales. The sample was based on a pilot study, and the general-school children were excluded if their ASRS full-scale T-score was over 60, which may affect how the non-case group represents all children.

Declared interests

The paper states the Chinese ASRS was developed by a team of the Research Special Fund for the Public Welfare Industry of Health of China. The supplied text gives no author conflict-of-interest statement.

The easy way to misread this

Do not read the converted score as better for diagnosis. The paper reports raw-score area under the curve 0.835 and converted-score area under the curve 0.834, and says there was no meaningful difference in diagnostic accuracy. Do not use the raw total score directly for screening or diagnosis; the paper says the ASRS T score should always be used for individual children.

Read it on PubMed →