Differential length and overlap with the stem in multiple-choice item options: A pilot experiment
Multiple-choice items are extensively used across different assessment contexts. A crucial requirement for ensuring their validity is their correct development, and a number of item-writing guidelines have been proposed that support item developers. This experimental pilot study aimed to investigate...
| Autores: | , |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2019 |
| País: | España |
| Institución: | Universidad Autónoma de Madrid |
| Repositorio: | Biblos-e Archivo. Repositorio Institucional de la UAM |
| Idioma: | inglés |
| OAI Identifier: | oai:repositorio.uam.es:10486/689866 |
| Acceso en línea: | http://hdl.handle.net/10486/689866 https://dx.doi.org/10.5093/psed2018a20 |
| Access Level: | acceso abierto |
| Palabra clave: | Flawed items Item difficulty Item-writing guidelines Length of options Lexical overlap Multiple-choice items Psicología |
| Sumario: | Multiple-choice items are extensively used across different assessment contexts. A crucial requirement for ensuring their validity is their correct development, and a number of item-writing guidelines have been proposed that support item developers. This experimental pilot study aimed to investigate the effect of violating two item-writing guidelines: the differential length of the correct option compared to distractors and its lexical overlap with the stem. Standard and flawed items, respectively adhering to and deviating from guidelines, were randomly assigned to 55 college students and compared in their psychometric functioning. Results indicated that, in general, flawed items tended to be easier and less subject to random answers than standard ones, but significant differences were few. Discrepancies between standard and flawed subtests approached statistical significance with medium effect sizes. Although of interest, findings must be cautiously interpreted due to the small sample size. Implications for future research are discussed. |
|---|