The valuation of academic achievements in students with severe language impairment is problematic if they also have difficulties in sustaining attention and in praxic skills. In severe autism all of these difficulties may occur together. Multiple-choice tests offer the advantage that simple praxic skills are required, allowing the tasks to be performed without physical support. Even so, attentive and behavioral difficulties may be so disruptive that achievements may be underestimated. Since special needs educators can give immediate feedback on each answer, a strategy might be to permit corrections, allowing further attempts, in order to mitigate these problems and to better capture their knowledge. Here a Microsoft Excel applet is designed to compute the statistical significance and the final grade of multiple-choice tests, if up to two corrections per selection are allowed. The method was used with a nonverbal student with severe autism and Down syndrome in a mainstream secondary school.