• https://publications.iadb.org/en/publication/17155/concurrent-validity-and-feasibility-short-tests-currently-used-measure-early

Rubio-Codina, Marta (2016). Replication Package for: Concurrent Validity of Short Tests for Early Childhood Development. IDB Open Data. https://doi.org/10.60966/n2y42g0v

1. English

What is the Bayley test used in this dataset?

The Bayley Scales of Infant and Toddler Development (Bayley-III) is a standardized assessment of early development. It measures Cognitive, Language, and Motor domains and reports composite scores that are normed with a mean of 100 and SD of 15. In this dataset, Bayley-III was administered alongside short tests (ASQ-3, Denver, Battelle, MacArthur) to examine concurrent validity in a Bogotá sample.

Who was assessed and when?

• Sample size: 1,311 children
• Location: Bogotá, Colombia (low and lower-middle income households)
• Age range: 6–42 months (median 24 months)
• Study design: Cross-sectional fieldwork linked to the paper on concurrent validity of short tests in large-scale studies (IDB, 2011 field data; dataset guide dated Aug 2016).

What are the average development scores (Colombia sample) on the Bayley test?

Across the full sample (n=1,311), Bayley-III composite score means were: - Cognitive: 98.4 (SD 8.9)
- Language: 96.5 (SD 9.9)
- Motor: 99.5 (SD 10.8)

Reminder: Bayley composite scores are scaled to mean=100, SD=15 in the reference norming sample. This Bogotá distribution reflects this specific study population and age mix.

Does the dataset include tests beyond the Bayley test?

Yes. Variables and scripts in the package indicate concurrent measures from: - ASQ-3 (Ages & Stages Questionnaires)
- Denver (Denver Developmental Screening)
- Battelle
- MacArthur (CDI)
These allow comparisons of short tests with the Bayley-III for concurrent validity analyses.

How should I interpret the Bayley composite scores?

• 100 is the norm mean; 15 points is one standard deviation.
• Scores 85–115 are within ~±1 SD of the norm mean.
• Scores below 85 suggest potential developmental delays and warrant clinical follow-up; scores ≥115 indicate higher-than-average performance. Interpretation must account for age, test conditions, language of administration, and socioeconomic context.

Are there citywide or national benchmarks to compare with these Bogotá results?

The dataset is designed for concurrent validity across instruments rather than for official city- or national-level benchmarks. For benchmarking, you would either: - Compare against Bayley-III norms (mean 100, SD 15), and/or
- Construct age-specific Bogotá reference tables from this sample with appropriate weights and confidence intervals.

What else is included to help researchers?

• A guide (PDF) describing the study, measures, and file organization.
• Do-files for correlation tables, bootstrapped standard errors, and tests of differences in concurrent validity.
• A consolidated data file (tests-bogota) with 311 variables, including Bayley domains, short-test scores, and rich child/household covariates.

What are the dataset’s limitations?

• Sample representation: The dataset focuses on a specific urban population in Bogotá, which may limit generalization to rural or national contexts.
• Age range: It covers a restricted age window (typically infants and toddlers), making it unsuitable for broader child or adolescent development analysis.
• Test adaptation: The short test versions were adapted for research feasibility; some psychometric precision is sacrificed compared to the full Bayley scale.
• Cross-sectional design: Because the data capture a single time point, the dataset cannot measure long-term developmental trajectories or causal effects.

What kinds of policy decisions can be informed by this dataset?

• Early childhood measurement design: Policymakers can use findings to validate shorter, cost-effective tools for national surveys, allowing broader data collection at lower cost.
• Targeting early interventions: Insights can guide early childhood development (ECD) programs to identify at-risk populations using efficient screening tools.
• Program evaluation: Results support better impact evaluation of early learning and parenting interventions by ensuring outcome measures are valid and scalable.
• Standardization of developmental metrics: The data contribute to regional efforts to harmonize ECD measurement standards across Latin America.

What is the Bayley Scales of Infant and Toddler Development?

The Bayley Scales are considered the “gold standard” diagnostic test for measuring developmental levels of infants and toddlers. They assess multiple domains including cognition, language, and motor skills, and are widely used in controlled research and clinical settings. In the Bogota study, the Bayley-III was administered by trained psychologists in standardized environments.
"Multi-dimensional diagnostic tests such as the Bayley Scales of Infant Development are considered to be the 'gold standard' to measure the developmental levels of infants and toddlers."

How long does a Bayley assessment take and what happens during it?

The Bayley-III was administered in public libraries or childcare centers, typically requiring a dedicated session with trained psychologists. Each child was tested individually in quiet, well-lit environments, with caregivers present. The dataset notes that Bayley-III administration is time-consuming and requires specialized training.
"Test administration is time consuming and requires highly trained professionals working in controlled environments."

Bayley Scales vs ASQ-3: which developmental screening is better?

The dataset finds that ASQ-3 performs poorly under 31 months, while Denver-II shows stronger validity. ASQ-3’s cognitive, language, and fine motor scales had low concurrent validity with Bayley-III below 19 months.
"Overall, the Denver-II seems the most feasible and valid multi-dimensional test and the ASQ-3 performs poorly under 31 months."

How reliable are the Bayley Scales for developmental delay, and what is their validity?

The dataset confirms Bayley-III as the gold standard with sensitivity to differences in outcomes due to interventions. It was used as the benchmark for concurrent validity comparisons.
"Importantly, this test has shown sensitivity to differences in ECD outcomes due to interventions in diverse contexts."