Autor(es): Khorramdel, Lale ; Yin, Liqun ; Foy, Pierre ; Jung, Ji Yoon; Bezirhan, Ummugul; von Davier, Matthias
Organisation(s): International Association for the Evaluation of Educational Achievement, IEA; UNESCO Institute for Statistics
Pages: 68 p.
This report is concerned with establishing a concordance between the regional ERCE and the international TIMSS and PIRLS achievement scales in Latin American and Caribbean countries. The Rosetta Stone study consists of two assessment parts. The first part is the ERCE assessment including the ERCE context questionnaire. The second part is the Rosetta Stone assessment comprising test booklets with item blocks and passages from TIMSS and PIRLS. Both assessment parts were administered in two ERCE countries to the same students on two consecutive days. Analyses were conducted using classical item statistics, item response theory (IRT), and population modeling. They comprise the evaluation of the data quality, evaluation of the psychometric quality of the instruments, establishing common scales across countries and assessments, and constructing concordance tables that account for the uncertainty of the measurement (measurement error).