Abstract
Objective: To develop a valid and reliable questionnaire to measure beliefs about autism in care staff working with people with learning disabilities. Procedure: Items for the Autism Beliefs Questionnaire (ABQ) were generated by: carrying out a thorough literature review, running a consultative meeting of clinical psychologists, getting experts in autism and an expert in psychometrics to make quantitative judgements about ABQ items, and conducting a pilot study with care staff. Then the ABQ was administered to care staff and experts, with test-retest data obtained 2-4 weeks later. Participants: Ninety-four care staff and 12 experts were administered the ABQ and a demographic questionnaire, and 60 care staff and 8 experts completed the second administration. Statistical analysis: Factor analysis was performed on the ABQ data. The relationship between individual ABQ items and demographic factors (type of service, experience of autism, and training in autism) were assessed using Kruskal-Wallis and Mann Whitney tests. The test-retest reliability of ABQ items was examined using Kendall’s tau correlations. Results: Factor analysis of the ABQ revealed no clear underlying factorial structure, even following removal of items. Twenty-seven ABQ items were found to have poor to acceptable test-retest reliability. On 14 items there were significant differences between experts and care staff (with differing levels of training in, and experience of, autism), demonstrating that the ABQ seemed to have reasonable concurrent validity. Finally, a revised version of the ABQ was produced for use as an evaluative tool for staff training and for research purposes.