Conditional Standard Error of Measurement
The conditional standard error of measurement (CSEM) describes how much measurement error a test score carries at each point along the score scale, rather than as a single average. A test typically measures more precisely in some score ranges than others — often best near the middle and worst at the extremes — and the CSEM captures that variation. Recognized in test theory by Lord and required by the Standards for Educational and Psychological Testing, it is essential for honest score reporting, especially near cut scores where classification decisions are made.
Kilderegistrering
Citater kopieret ordret fra metodens kilderegistrering. Ingen påstandsniveauverifikation er udledt heraf.
- Lord, F. M. (1980). Applications of Item Response Theory to Practical Testing Problems. Lawrence Erlbaum Associates. · ISBN 9780898590067
- American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. (2014). Standards for Educational and Psychological Testing. AERA. · ISBN 9780935302356
Kuraterede påstande
Påstande gemt i bevis-loggen, hver med sin egen vurdering.
Denne visning opfinder ikke en påstandsvurdering, når loggen ingen har.
Relaterede metoder
Genereret fra metodegrafen og vist som maskinelt foreslåede relationer — ingen bevispåstand er udledt.