Re-Evaluating the Appropriateness of the "Don't Know" Response Option: Guessing Rate as a Source of Systematic Error on Autism Knowledge Assessments.

Participants were recruited through professional organizations, so the sample may not represent all school-based professionals. Only audiologists, speech-language pathologists, and school psychologists who worked mainly in school settings were included. 171 of 567 initial participants were excluded. 75 did not complete any autism knowledge questions, and 19 had missing predictor data. The final sample was 396 participants; 363 were female, 334 had a master’s or specialist’s degree, and 324 personally knew someone with autism. The study used only two autism knowledge measures, ASKSP-R and ASK-Q, and other measures with a “don’t know” option may behave differently. Personal autism contact was recorded as yes or no, and the number of autistic students worked with was a self-reported ordinal estimate. The study did not ask whether participants themselves were autistic. This was a survey analysis, so it shows links between guessing and scores, not proof that guessing causes higher knowledge scores.

Do not read a low autism knowledge score as proof that a professional lacks autism knowledge. In this study, don’t know or skipped answers were marked incorrect in one analysis, and randomly assigning those answers raised average scores. The finding is about test format and guessing, not about patient care.