Observation is a valid way of assessing common variables in typical babbling and identifies infants who need further support.

Aim: It is important to identify speech and language difficulties in children as early as possible. This study investigated the validity of observing babbling structure and consonant articulation.

Methods: We focused on the language capabilities of children, with and without cleft palates, at 12 months of age (n = 29) and 18 months of age (n = 38), comparing observations made by speech and language pathologists on babbling structure and consonant articulation to audio recordings phonetically transcribed by experts blinded to the aim of the study. Descriptive cross-tabs, sensitivity, specificity and predictive values were analysed to investigate the validity of agreement between the methods.

Results: We found high mean agreement between the two methods (94%) and within the observation (100%) and transcription (88.5%) teams when it came to canonical babbling, high-pressure/oral stop consonants and anterior placement of bilabial and dental/alveolar consonants. The observations had high specificity and negative predictive values (0.90-1) at both ages, with low to moderate sensitivity (0-0.86). However, there was low agreement, between and within the groups, on the number of consonant types and the unproven predictive variable glottal place of articulation.

Conclusion: Previously suggested predictive measures of babbling were shown to be valid and observation is a reliable method of identifying children who need further support.