The P-Gsq: A New Self-Report Sensory Questionnaire for Children

Background:   The Glasgow Sensory Questionnaire (GSQ; Robertson & Simmons, 2013) is a 42-item self-report questionnaire which measures the frequency and severity of sensory difficulties (both hyper- and hypo-responsiveness), giving a score ranging from 0 (no sensory difficulties) to 168 (Extremely severe and frequent difficulties in all sensory modalities). Research with this questionnaire has shown that there are strong correlations of sensory score with self-reported autistic traits (Robertson & Simmons, 2013; Horder et al, 2014) and (in a Japanese translation) significant differences in sensory scores between diagnosed ASD and control populations (Takayama et al, 2014). However, so far the GSQ has only been critically tested with adults. We have therefore developed a version of the questionnaire adapted for children: the Pictured-Glasgow Sensory Questionnaire (P-GSQ).

Objectives:  To develop and validate a version of the GSQ adapted for children (8-15 years).

Methods:   The questionnaire adaptation was achieved by (a) modifying the language slightly to make it understandable to a typical 8-year-old and (b) incorporating eye-catching illustrations to support the question text. After performing pilot experiments to verify that the text and pictures were adequately understood, the questionnaire was completed by 234 children (mean age 10.15 years, age range 8-15 years) during the course of a public workshop at a local “hands-on” science museum. The caregiver with the child was asked to complete the Children’s version of the AQ (Auyeung et al, 2009) at the same time, so that the link between P-GSQ and autistic trait level could be evaluated.

Results: Overall results show only a modest, if significant, correlation between P-GSQ and AQ score (Pearson’s r(234) = 0.147, p < 0.05, R² = 0.022). Subsequent analysis of sub-scales of the P-GSQ indicated that this correlation was largely driven by auditory and gustatory hyper-sensitivity and by the proprioceptive modality. Splitting the ages up into two groups showed that the correlation was only significant in the 8-11 age range and not in the 12-15-year-olds. Splitting the group by gender found that the correlation was only significant in females and not males. Cronbach’s Alpha revealed a good degree of internal consistency (alpha = 0.835). 

Conclusions:   These results are curious given the well-documented and replicated significant correlations between GSQ sensory scores and autistic trait levels found in adults (16+). Pilot data collected with 12-15-year-olds and the adult GSQ in a similar way to these data also revealed no significant correlations. The general response seems to be a modest amount of sensory difficulty (somewhere between “often” and “sometimes”) for most of the young people we tested. One explanation is that all young people have mild difficulties with sensory stimulation: the difference is how they react to it, and this “sensory reactivity” is what is picked up in the more-frequently used caregiver questionnaires. Alternatively, young people may lack the insight necessary to self-report on their own sensory difficulties and/or the difficulties that they experience may be qualitatively different from those that adults experience.