Validation of a Mathematics Test Applied to Students Starting Higher Education, Using the Rasch Model
In this test, the Rasch model was applied for the calibration of a valid and reliable assessment instrument. Between the periods 2020-I to 2022-I the diagnostic test consisted of a 20-item mathematics questionnaire taken prior to a level zero course at “Instituto Superior Universitario Central Técni...
| Autores: | , , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2023 |
| País: | Ecuador |
| Institución: | Universidad Central del Ecuador |
| Repositorio: | Revista Ingenio |
| Idioma: | español |
| OAI Identifier: | oai:revistadigital.uce.edu.ec:article/4548 |
| Acceso en línea: | https://revistadigital.uce.edu.ec/index.php/INGENIO/article/view/4548 |
| Access Level: | acceso abierto |
| Palabra clave: | Modelo matemático de Rasch Teoría de Respuesta al Ítem habilidad dificultad distribución estadística Rasch mathematical model Item Response Theory ability difficulty statistical distributions |
| Sumario: | In this test, the Rasch model was applied for the calibration of a valid and reliable assessment instrument. Between the periods 2020-I to 2022-I the diagnostic test consisted of a 20-item mathematics questionnaire taken prior to a level zero course at “Instituto Superior Universitario Central Técnico”. With this test, 695 students had been evaluated in the periods 2020-I, 2020-II, 2021-I and 2021-II; subsequently, the items of the evaluation instrument that were not described by the Rasch model were identified with a reliability of 65%, which were corrected. Finally, with the new corrected test, another 100 students had been evaluated in the period 2022-I, obtaining a test reliability of 90%. From these results, the characteristic curves of these items were generated, applying Pearson and chi-square distributions, those that did not fit the model were identified. Using the parameters obtained by the Rasch model, the grades were simulated and compared with the actual grades obtained by the students. Thus, the model has made it possible to identify 133 students with a low level of ability, of which 119 correspond to the original test and 14 to the corrected test. R software was used for the statistical analysis. |
|---|