Reliability of speech diadochokinetic test measurement

Background: Measures of articulatory diadochokinesis (DDK) are widely used in the assessment of motor speech disorders and they play a role in detecting abnormality, monitoring speech performance changes and classifying syndromes. Although in clinical practice DDK is generally measured perceptually, without support from instrumental methods that display the acoustic waveform, no standard measurement procedures are employed and studies on inter- and intra-rater reliability are lacking. Aims: To investigate the inter- and intra-rater reliability of perceptual DDK measurement and the possible impact of experience on rating performance. Methods Procedures: Ten speech and language therapists and ten untrained controls rated 12 DDK speech samples, including two pairs of identical samples, of speakers with different neurological speech disorders. We compared timing for repetitions as close as possible to 5s, for 5s exactly, and for as long as the speaker could manage; as well as four qualitative features (rhythm, distinctness, loudness and overall impression). Counts by time were compared with objective outcomes from sound spectrograms. Outcomes Results: Inter- and intra-rater reliability was lower than should be acceptable for clinical diagnostic and outcome assessment. The most accurate and repeatable procedure (compared with objective measurement) was counting the syllables within the first 5s. There was no unequivocal influence of clinical experience on accuracy. Conclusions: The findings suggest caution in employing and interpreting clinical DDK measures as diagnostic and outcome measures. More research is necessary to settle the influence of experience on time and count accuracy and what modifications to timing might bring increased agreement that permits DDK to realize its full potential in assessment.