Construct Validity of the Viking Speech Scale.

Small secondary data sets collected for other studies, not a new validation trial designed for this purpose. The U.K. and U.S. samples differed in age, dysarthria presence, and speech severity, so findings are not one homogeneous group. Intelligibility was measured differently in the two countries: repetition and transcription in the U.S., forced-choice single words and spontaneous picture-description connected speech in the U.K. There were very few children rated at the extreme VSS levels, with level I only in the U.S. sample and level IV only in the U.K. connected-speech data. Intelligibility scores overlapped within VSS levels, especially levels II and III, so a VSS level does not map to one exact intelligibility score. Mean intelligibility was based on three listeners per child, so one listener's score could shift a child's average. No reliability checks were completed for the descriptor measures in these data subsets. The small samples did not allow testing of which child characteristics, such as dyskinesia, influence VSS rating.

The declaration states that this was independent research funded in part by the National Institute for Health Research under its Research for Patient Benefit Programme.

Do not treat a Viking Speech Scale level as a precise intelligibility score. The study found that intelligibility predicted VSS level, but scores overlapped within levels, especially levels II and III, and the VSS was rated from therapist observation rather than only from audio intelligibility.