| |  Verfügbarkeit |
| Autoren | Fischer, Luise; Rohm, Theresa; Gnambs, Timo; Carstensen, Claus H. |
|---|
| Institution | Leibniz-Institut für Bildungsverläufe; Nationales Bildungspanel |
|---|
| Titel | Linking the data of the competence tests. |
|---|
| URL | https://www.neps-data.de/Portals/0/Survey Papers/SP_I.pdf |
|---|
| Erscheinungsjahr | 2016 |
|---|
| Seitenzahl | 20 S. |
|---|
| Verlag | Bamberg: Leibniz Institute for Educational Trajectories (LIfBi) |
|---|
| Reihe | NEPS survey paper. Band 1 |
|---|
| Dokumenttyp | Monographie; Graue Literatur; online |
|---|
| Beigaben | Literaturangaben, Tabellen, Anhang |
|---|
| Sprache | englisch |
|---|
| Forschungsschwerpunkt | Bildungspanel (NEPS) |
|---|
| Schlagwörter | Kompetenzmessung; Kompetenzentwicklung; Skala; Datenverarbeitung; Lesekompetenz; Schuljahr 05; Schuljar 07 |
|---|
| Abstract | The National Educational Panel Study (NEPS) aims at investigating the development of competencies across the whole life span and developing tests for assessing different competence domains. In order to compare competencies across different measurement occasions and examine competence development over time the different measurements must be placed onto a common scale. This goal is achieved by linking two measurements of the same construct. The present document describes the linking procedure adopted for the NEPS. Moreover, this approach is demonstrated using data from Starting Cohort 3 that links mathematical and reading competences across Grades 5 and 7. First, the procedure on how to derive linked item parameters in an anchor-items design is described. After showing that the mathematical tests administered in Grades 5 and 7 are unidimensional, it is demonstrated that all common items administered in both grades showed negligible differential item functioning. Therefore, the two mathematical tests were linked using sixmeasurement invariant items. Subsequently, the procedure on how to derive linked item-parameters in an anchorgroups design is described. It is demonstrated that the reading competence tests administered in Grades 5 and 7 were unidimensional and showed no differential item functioning. Therefore, the two reading tests were linked using responses from an independent link sample. Finally, the naming conventions in the scientific use file for repeatedly administered items and different competence scores are presented. (Orig.). |
|---|
| Förderkennzeichen | 01GJ0888 |
|---|
