Psychometric Properties of the Spence Children's Anxiety Scale: Parent Report in Children with Autism Spectrum Disorder.

Jitlina et al. (2017) checked if the full Spence Children’s Anxiety Scale–Parent Report (SCAS-P) works in autistic children. They ran a factor analysis on parent answers from a large group of kids with autism. The goal was to see which parts of the scale keep their meaning in this population.

The six-factor structure of the full SCAS-P did not hold up. Only four subscales—Generalized Anxiety, Separation Anxiety, Panic, and Agoraphobia—kept good reliability and validity. The other two subscales (Social Anxiety and Physical-Injury Fears) fell apart, so the authors advise skipping them.

Magiati et al. (2017) ran almost the same study the same year and reached a similar conclusion: the total SCAS-P score is reliable for screening, but the six-factor model is messy. Together, the two papers act like a direct replication, boosting confidence that clinicians should use the total score or selected subscales, not the full factor map.

Chowdhury et al. (2016) and Tde Wit et al. (2024) echo the same recipe: start with a long parent questionnaire, run a factor analysis, and keep only the cleanest items. The HSQ-ASD and ASDQ both ended up shorter and stronger, just like the trimmed SCAS-P.

Green et al. (2020) and Saad (2025) show the same pattern in stress scales: full versions wobble, but brief, well-fitting subscales survive. The message across studies is consistent—shorter, targeted tools outperform long ones in autism samples.

When you assess anxiety in autistic clients, grab only the four solid SCAS-P subscales. You will save parents time, cut scoring errors, and still track the symptoms that matter. Drop the Social Anxiety and Physical-Injury items unless you want noisy data. This pick-and-choose approach is now backed by two independent 2017 studies and mirrors the shortening strategy that works for other parent scales.