The Social Communication Questionnaire (SCQ) is a 40-item yes/no questionnaire completed by a parent or primary carer to screen for autism spectrum conditions. Its items were drawn from the Autism Diagnostic Interview-Revised (ADI-R) and sample the three classic domains of that interview: reciprocal social interaction, language and communication, and restricted, repetitive and stereotyped patterns of behaviour. It is intended for anyone aged four or over with a mental age of at least two, and takes a carer under ten minutes to complete without supervision.
Two forms exist. The Lifetime form asks about the whole developmental history, with particular attention to behaviour between the ages of four and five, and is the form validated for screening. The Current form asks about the past three months and is offered for treatment planning and monitoring. The SCQ is a screen, not a diagnostic instrument: a positive result indicates that a fuller assessment such as the ADOS-2 or ADI-R is warranted, and a negative result lowers but does not exclude the probability of autism.
02 - Origin & purpose
Where it comes from.
The instrument was developed by Sibel Berument, Michael Rutter, Catherine Lord, Andrew Pickles and Anthony Bailey and published in the British Journal of Psychiatry in 1999 as the Autism Screening Questionnaire. Its authors wanted a brief, parent-completed companion to the ADI-R that could be used to identify who should receive the full interview. The initial validation compared 160 individuals with pervasive developmental disorders against 40 with other diagnoses, and found that a cutoff of 15 best separated the two groups.
Western Psychological Services published the instrument in 2003 under the name Social Communication Questionnaire, with Rutter, Bailey and Lord as authors, partly to avoid confusion with the Ages and Stages Questionnaires. Since then it has been translated into many languages, used to characterise research cohorts, and adopted as a first-stage screen in developmental and mental-health services. Three subscales corresponding to the ADI-R domains have been described, but they are not part of the standard scoring forms and have limited supporting research.
03 - Scoring & cutoffs
How scoring works.
Each item is scored 0 or 1 in the direction of autistic behaviour. The first item asks whether the child speaks in phrases; if so, items 2 to 40 are summed (39 items, range 0 to 39), and if not, items 8 to 40 are summed (33 items), because the language items cannot be scored in a non-verbal child. The published cutoff of 15 or more indicates possible autism spectrum disorder and warrants referral for diagnostic assessment; a score of 22 or more was proposed to distinguish autism from other spectrum conditions, although this second threshold discriminates less well.
Subsequent studies have repeatedly found that 15 is too conservative in young children. In the CDC's Study to Explore Early Development, preschool children scored on the Current form reached sensitivity of 87% with specificity of 81% at a cutoff of 11, 75% and 86% at 13, and only 63% with 91% specificity at 15. Corsello and colleagues found sensitivity and specificity of 0.71 at 15 across ages two to sixteen, with lower sensitivity under eight, and suggested a cutoff of 12 for children under four. Services therefore often use 11 to 13 in young or developmentally delayed children and accept more false positives in exchange for fewer missed cases.
04 - Validation evidence
How well it performs.
The SCQ has a large validation literature, summarised in a 2017 meta-analysis by Chesnut and colleagues that pooled an area under the curve of 0.885 across studies. Accuracy is highest with the Lifetime form in children aged four and over drawn from clinical or research referral samples: Chandler and colleagues found an area under the curve of 0.88 for spectrum versus non-spectrum cases in nine- and ten-year-olds, and 0.93 for autism versus non-autism at the 22 cutoff. Internal consistency is reported as 0.87 in the manual and 0.89 to 0.94 in later samples, with test-retest intraclass correlations of 0.87 to 0.96 in a Turkish validation. Correlation with the ADI-R total is high (r = 0.83 in Charman et al. 2007), as expected given the items were derived from it. Performance falls sharply in children under four, in intellectual disability, and in community mental-health samples where comorbidity is high.
85% / 75%
Screening accuracy (ASD vs non-ASD)
Sensitivity / specificity at the 15 cutoff in the original validation - Berument et al. 1999
AUC 0.885
Pooled diagnostic accuracy
Meta-analysis across validation studies; falls to 0.77 under age 4 - Chesnut et al. 2017
90% / 86%
Autism vs non-autism
Sensitivity / specificity at the 22 cutoff, AUC 0.93, age 9-10 - Chandler et al. 2007
α 0.87-0.94
Internal consistency
0.87 in the WPS manual; 0.89 (non-verbal) to 0.94 (verbal) in Marvin et al. 2017
05 - How it compares
How it compares to the alternatives.
Instrument
Items
Time
When to reach for it
SCQ
40 yes/no, carer-report
<10 min
Licensed (WPS/Hogrefe). Reach for it as a first-stage carer screen from age 4 when full ADOS-2/ADI-R assessment is being rationed, or in research cohorts anchored to the ADI-R.
Free for non-profit use. Reach for it when the person can self-report - adolescents and adults - and you need a quick NICE-recommended referral screen.
M-CHAT-R/F
20 + follow-up interview
5 min + follow-up
Free for non-profit clinical use. Reach for it in toddlers aged 16-30 months, where the SCQ has no validity.
CAST
39, carer-report
~10 min
Free from the Cambridge Autism Research Centre. Reach for it as a no-cost carer screen for children aged 4-11 in mainstream settings.
SRS-2
65 per form
15-20 min
Licensed (WPS). Reach for it when you need a dimensional severity score across social responsiveness rather than a yes/no screen, or to track change.
ADOS-2
Structured observation
40-60 min
Licensed (WPS) with training. Reach for it as the diagnostic observation once a screen is positive - it is not a screening tool.
06 - When to use it
Right tool, wrong tool.
Reach for it when
-Children and adults aged 4 and over (mental age 2 and over) where a parent or carer who knows the developmental history can complete it.
-Services holding a WPS or Hogrefe licence that need a standardised first-stage screen before booking ADOS-2 or ADI-R time.
-Research cohorts that need an ADI-R-anchored screen or a quick check of case status.
-The Current form for describing present-day behaviour in someone already diagnosed, provided results are interpreted descriptively.
Reach for something else when
-Children under 4, and toddlers in particular - use the M-CHAT-R/F, which is designed for 16-30 months.
-Adolescents or adults presenting themselves, without a carer - use the AQ-10 as a self-report screen.
-Community CAMHS caseloads with mixed anxiety, ADHD and mood comorbidity, where specificity at the 15 cutoff has been as low as 12-26%.
-Intellectual disability without an adjusted cutoff - specificity drops considerably.
-Any use as a diagnosis, or to rule autism out on its own.
-Services without a licence - use the CAST or AQ-10 instead.
07 - Confidence & precision
Reading the score with care.
No standard error of measurement or minimal clinically important difference has been published for the SCQ, and because the Lifetime form asks about developmental history it is not designed to measure change at all. The Current form is offered for monitoring, but its responsiveness has not been formally studied and using it for screening lowers accuracy substantially. Test-retest intraclass correlations of 0.87 to 0.96 imply reasonable stability, but scores within a few points of the cutoff should be treated as borderline rather than decisive: in general-population samples around 4 to 5% of children score at or above 15, and diagnosed girls score on average around four points higher than diagnosed boys, suggesting that girls near the threshold are more likely to be missed.
08 - Limitations
What it cannot tell you.
The SCQ performs poorly under age four, with an area under the curve of 0.77 in the pooled analysis and sensitivity as low as 56% in two-year-olds. Specificity falls in intellectual disability, in children with high levels of behaviour problems, and in families with lower socioeconomic status or maternal education. In community child mental-health samples, where referrals arrive with mixed diagnoses, discrimination has been close to chance (AUC 0.52 in Hollocks et al. 2019). Its items were derived from a predominantly male phenotype, so more subtly presenting girls may be under-identified. Using the Current rather than Lifetime form for screening reduces accuracy markedly. Finally, it is a licensed instrument that must be purchased, cannot be adapted or hosted online without WPS permission, and requires Level C qualification.
09 - Licensing, explained
How licensing works.
The SCQ is copyrighted and sold by Western Psychological Services (Hogrefe Ltd in the UK) at Level C qualification. WPS does not permit its items, scoring or manual to be reproduced, so this page explains how the SCQ works without showing it. For a free autism screen you can score here, use the AQ-10.
See how Aisel removes friction where it costs most. A 20-minute walkthrough tailored to your clinic.
We value your privacy
We use cookies to analyse site usage and improve your experience. Analytics and embedded media (e.g. YouTube) only load if you accept. Read our cookie policy.