TALIS Starting Strong 2018 Data
- All parameters estimated were weighted to generate reliable, valid, and comparable estimates of population parameters based on sample data.
- Center weights were used to weight leader data, while staff weights were used to weight staff data or combined staff and center leader data.
- Standard errors were estimated by applying Fay’s variant of the balanced repeated replication (BRR) technique.
- Percentiles of each national sample were categorized into four “national quarters,” by first sorting and creating a vector of accumulated weighted frequencies, and then allocating them to their respective national quarter.
Suggestions for doing secondary analyses:
All analyses in TALIS Starting Strong 2018 should be computed using the International Database (IDB) Analyzer latest version or comparable software that can account for the complex sample design of TALIS Starting Strong. Population weights and BRR replicate weights for variance estimations must be used.
Data archives have been created and are made available in SPSS, SAS, and STATA format.
Data files exist for staff and leaders for the two populations: ISCED 02 and settings for children under age three.
Data files are available on the OECD TALIS Starting Strong website for each participating country. Combined international public leader and staff files are also available.
All public-use data files are available for download at the OECD TALIS Starting Strong website.
Special permission is needed to obtain data from Iceland.