Loading...
Loading...
Contributors:
Dataset URI:
10.5281/zenodo.23499187Derived data files, with the scripts that produce every one of them, for the 69 education systems with at least 500 student responses to the PISA school belonging items (ST034) in both PISA 2018 and PISA 2022. For each system the atlas reports whether the belonging scale measures the same thing in the two cycles (configural, threshold, metric and scalar invariance), which item is responsible where it does not, and the change in the latent means once comparability is established. It also maps which questionnaire items can be compared at all, documents the rotated questionnaire design of 2022, and reports socioeconomic gradients by cycle. Comparability here is over time within each system, not across systems. Main results. The six items fit two factors separated by wording (CFI 0.997, RMSEA 0.041; one factor: 0.906 and 0.214). With WLSMV and the conventional cut-offs, scalar invariance holds in 65 systems and fails in France, Macao, Montenegro and Viet Nam. Freeing one intercept restores comparability in the first three; Viet Nam fails already at the threshold step, needs two freed items, and its estimates are not recommended. With a stricter cut-off, 20 systems fail, including all 7 atlas systems to which the OECD gave version-specific item parameters. Changes go in both directions and are mostly small (median absolute change 0.078 and 0.069 standard deviations on the two factors), with replicate-based standard errors, and in the invariant systems they are closely correlated with the OECD index (r = 0.84 and 0.93). Where to start: data/02_invariance/belonging_invariance_by_country.csv (is the comparison defensible for a system?), data/02_invariance/belonging_partial_invariance.csv (which item, and is partial invariance enough?) and data/03_latent_means/latent_means_master.csv (the recommended file for the size of the change). The README, the change log and a data dictionary in Markdown, JSON and CSV describe every file and column. Version 2.1.0 adds, without changing any existing value, an alignment check of the latent change in the four systems estimated under partial invariance and three reference systems (data/03_latent_means/alignment_sensitivity.csv, sirt 4.2-133, items treated as continuous), the two figures of the data descriptor, and every CSV file deposited individually next to the zip; it lists the four authors of the data descriptor. Version 2.0.0 replaced version 1.0.0 and accompanied the first revision of the data descriptor submitted to Data (MDPI). It renames every file and column in English (with a map from the old names), adds the data dictionary and dependency files, estimates all measurement models with the final student weight and the standard errors of the latent changes with the 80 replicate weights, adds partial-invariance and sensitivity files, and corrects the reading of the direction of the factors: the items are coded from 1 (strongly agree) to 4 (strongly disagree), so the column diff_belonging gives both factors in the direction of more belonging. See docs/CHANGELOG.md. This package does not contain PISA microdata. The sources are the PISA 2018 and PISA 2022 public-use files distributed by the OECD. PISA 2025 is not covered.
Keyword(s):
alignment; cross-cycle comparability; large-scale assessment; measurement invariance; open data.
Grants and Funding:
No Grants or Funding Data Available
Associated Publication(s):
No Associated Publication Data Available
Supplementary Data:
No Supplementary Data Available
Related Dataset(s):
No Related Dataset Data Available
The National Library of Medicine (NLM) provides information about biomedical datasets. Inclusion on an NLM Web page does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: NLM Web Policies (nih.gov)