Theodora Papastefanou, Kakia Petinou
Speech sound disorders (SSD) are highly prevalent in preschool-aged children, yet their identification remains challenging in linguistically diverse contexts where speech variability may reflect dialectal difference rather than disorder. The present study investigated whether token-to-token speech inconsistency differentiates dialectal Cypriot-Greek-speaking children with typical development (TD) from peers with SSD. Twenty preschoolers (mean age = 5;6) participated: 10 with TD and 10 with SSD. Speech inconsistency was assessed using the DO.KI.GLO Inconsistency Protocol, a Cypriot-Greek adaptation of the DEAP inconsistency subtest, eliciting three productions of 25 familiar lexical items. Inconsistency was operationalised as phonological variation across repeated tokens and quantified using an Inconsistency Index. Results revealed a substantial group difference: children with SSD produced significantly higher inconsistency scores than TD peers (71.3% vs. 2.3%), with a large effect size. Nine of ten children in the SSD group exceeded the 40% threshold associated with Inconsistent Phonological Disorder. Qualitative analyses further differentiated groups, with TD variability reflecting predictable developmental substitutions, while SSD variability involved multi-feature, unpredictable substitutions, omissions, and distortions. These preliminary findings provide initial evidence supporting the clinical relevance of speech inconsistency in a dialectal context and highlight the importance of combining quantitative and qualitative analyses within culturally and linguistically sensitive assessment protocols for accurate SSD identification and subtyping.