Artificial (in)stability in school value-added estimates — poster, lightning talk and abstract, RSSB 2026

Poster, one-minute lightning-talk slide and abstract presented at the 33rd Annual Meeting of the Royal Statistical Society of Belgium (Antwerp, 19–20 November 2026). Cross-cohort correlations of estimated school value-added effects are the latent stability filtered through how much data we have, who is observed, and which school is credited. Three simulation studies isolate each factor at true ρ = .80: (1) information quantity and model choice, (2) selective missingness, (3) pupil mobility and attribution. The recurring finding: the joint model recovers the correlation while individual schools can still be misidentified. Stability and school identification are distinct inferential targets. Tom Van Ransbeeck (KU Leuven), with Koen Aesaert, Wim Van den Noortgate (KU Leuven) and George Leckie (University of Bristol). Contact: [email protected]. Study 1 is under journal review; preprints of Studies 2 and 3 will be added when available. Context and related work This dissertation belongs to the research line on the reliability and stability of learning-gain and value-added models of Koen Aesaert at the Centre for Educational Effectiveness and Evaluation (CO&E), KU Leuven.CO&E: https://ppw.kuleuven.be/onderwijskunde/en/team/team-coeKoen Aesaert: https://www.kuleuven.be/wieiswie/en/person/00119599 Related projects at CO&E:– Schoolresultaten door de lens van leerwinst: naar valide en bruikbare indicatoren (Schools’ learning gain in Flanders: towards valid and usable indicators). OBPWO policy research, Flemish Department of Education and Training, December 2025 – November 2027; promotor Koen Aesaert. https://research.kuleuven.be/portal/en/project/3H250609– De Vlaamse Toetsen: van resultaten naar leerwinst via twee meetpunten (The Flemish national tests: from results to learning gain via two measurement points). OBPWO 25.03, February 2026 – February 2030; co-promotor Koen Aesaert. https://research.kuleuven.be/portal/nl/project/3H260026 Dissertation project: The role of sample characteristics in the stability of school value-added estimates: sample size, student mobility and sample heterogeneity (registered at KU Leuven as “De rol van steekproefkenmerken in de stabiliteit van de schattingen van de toegevoegde waarde van scholen: steekproefgrootte, leerlingenmobiliteit en steekproefheterogeniteit”). KU Leuven, 2023–; promotor Koen Aesaert.

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-04
DOI
https://doi.org/10.5281/zenodo.23958295
Primary Topic
Educational Assessment and Improvement
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Artificial (in)stability in school value-added estimates — poster, lightning talk and abstract, RSSB 2026

Wim Van Den Noortgate, Koen Aesaert, T Van Ransbeeck, George Leckie
Zenodo (CERN European Organization for Nuclear Research)
Educational Assessment and Improvement
article

Artificial (in)stability in school value-added estimates — poster, lightning talk and abstract, RSSB 2026

Wim Van Den Noortgate, Koen Aesaert, T Van Ransbeeck, George Leckie
article en

Abstract

Poster, one-minute lightning-talk slide and abstract presented at the 33rd Annual Meeting of the Royal Statistical Society of Belgium (Antwerp, 19–20 November 2026). Cross-cohort correlations of estimated school value-added effects are the latent stability filtered through how much data we have, who is observed, and which school is credited. Three simulation studies isolate each factor at true ρ = .80: (1) information quantity and model choice, (2) selective missingness, (3) pupil mobility and attribution. The recurring finding: the joint model recovers the correlation while individual schools can still be misidentified. Stability and school identification are distinct inferential targets. Tom Van Ransbeeck (KU Leuven), with Koen Aesaert, Wim Van den Noortgate (KU Leuven) and George Leckie (University of Bristol). Contact: [email protected]. Study 1 is under journal review; preprints of Studies 2 and 3 will be added when available. Context and related work This dissertation belongs to the research line on the reliability and stability of learning-gain and value-added models of Koen Aesaert at the Centre for Educational Effectiveness and Evaluation (CO&E), KU Leuven.CO&E: https://ppw.kuleuven.be/onderwijskunde/en/team/team-coeKoen Aesaert: https://www.kuleuven.be/wieiswie/en/person/00119599 Related projects at CO&E:– Schoolresultaten door de lens van leerwinst: naar valide en bruikbare indicatoren (Schools’ learning gain in Flanders: towards valid and usable indicators). OBPWO policy research, Flemish Department of Education and Training, December 2025 – November 2027; promotor Koen Aesaert. https://research.kuleuven.be/portal/en/project/3H250609– De Vlaamse Toetsen: van resultaten naar leerwinst via twee meetpunten (The Flemish national tests: from results to learning gain via two measurement points). OBPWO 25.03, February 2026 – February 2030; co-promotor Koen Aesaert. https://research.kuleuven.be/portal/nl/project/3H260026 Dissertation project: The role of sample characteristics in the stability of school value-added estimates: sample size, student mobility and sample heterogeneity (registered at KU Leuven as “De rol van steekproefkenmerken in de stabiliteit van de schattingen van de toegevoegde waarde van scholen: steekproefgrootte, leerlingenmobiliteit en steekproefheterogeniteit”). KU Leuven, 2023–; promotor Koen Aesaert.

Zenodo (CERN European Organization for Nuclear Research)
University of Bristol (GB), IMEC (BE), KU Leuven (BE)
Openalex Percentile: Top 10%
Educational Assessment and Improvement
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon