Related Experiment Video
Updated: May 29, 2026
Behavioral Assessment of Hearing in 2 to 4 Year-old Children: A Two-interval, Observer-based Procedure Using Conditioned Play-based Responses
Published on: January 23, 2017
How to measure the onset of babbling reliably?
Inge Molemans1, Renate van den Berg, Lieve van Severen
1University of Antwerp.
Insights
Determining the start of babbling in infants is challenging. This study introduces a reliable bootstrapping method to accurately identify babbling onset using spontaneous speech data.
Area of Science:
- Developmental linguistics
- Child language acquisition
- Speech development
Background:
- Identifying the onset of babbling is crucial for understanding early language development.
- Existing measures lack formal definitions and sample size validation.
- Reliability issues arise from single computations and unequal sample sizes across infants.
Purpose of the Study:
- To assess the reliability of five common babbling onset measures.
- To propose and validate a robust procedure for identifying babbling onset with high confidence.
- To determine the minimal sample size required for reliable babbling onset detection.
Main Methods:
- Analysis of a large longitudinal corpus of spontaneous speech from 40 infants (aged 6-24 months).
- Experiment 1: Assessed reliability of five measures under varying conditions (single computation, unequal sample sizes).
- Experiment 2: Developed and validated a bootstrapping procedure for babbling onset identification.
Main Results:
- Establishing babbling onset with 95% confidence is unreliable with single computations and unequal sample sizes.
- Different measures require distinct minimal sample sizes for reliability.
- The proposed bootstrapping procedure reliably identifies babbling onset with 95% confidence.
- Bootstrapping requires relatively few data through extensive resampling.
Conclusions:
- Standard measures for babbling onset lack reliability and formal validation.
- A robust bootstrapping procedure offers a statistically sound method for identifying babbling onset.
- This validated procedure aids in accurate tracking of early language development milestones.
Abstract:
Various measures for identifying the onset of babbling have been proposed in the literature, but a formal definition of the exact procedure and a thorough validation of the sample size required for reliably establishing babbling onset is lacking. In this paper the reliability of five commonly used measures is assessed using a large longitudinal corpus of spontaneous speech from forty infants (age 0 ; 6-2 ; 0). In a first experiment it is shown that establishing the onset of babbling with reasonable (95%) confidence is impossible when the measures are computed only once, and when the number of vocalizations are not equal for all children at all ages. In addition, each measure requires a different minimal sample size. In the second experiment a robust procedure is proposed and formally defined that permits the identification of the onset of babbling with 95% confidence. The bootstrapping procedure involves extensive resampling and requires relatively few data.
