Huibin Zhang

Given the rapid growth in educational policies targeting educators' knowledge of dyslexia, this study explored the technical adequacy of a common instrument for measuring that knowledge. The responses of 1141 preservice teachers were scored in three ways: polytomously with the original 4‐point Likert scale, dichotomously as true‐false, and dichotomously as though the options were multiple choice. An exploratory factor analysis suggested at least one‐third of the items needed to be removed. Confirmatory factor analyses suggested a one‐factor model with polytomous scoring had the best fit to the data, but only six items loaded. All models demonstrated unacceptable internal consistency reliability (<0.70). Because no technically adequate version of the measure was identified, questions remain about basing policy on scores from these instruments. However, the findings indicated ways this type of measure might be improved.