Monitoring change requires a rethink of assessment practices in voice and speech

Abstract

Assessment for the purpose of monitoring change over time requires a different practical and statistical approach to that of assessment for diagnosing impairment. Sophisticated methods exist for identifying strengths and weaknesses in a patient's voice/speech profile, yet our understanding of the impact of repeated assessment is limited. Monitoring change necessitates that stimuli are stable in the absence of any true change in functioning, while remaining sensitive to influences that are considered to alter functioning (degeneration, therapy). The current paper discusses the issues relating to stimuli selection, identifying error within the sample and appropriate statistical models for identifying intra-individual change in the context of clinical and experimental speech or voice examinations.

Keywords: acoustics; repeated assessment; speech; statistics; voice.