The aim was to conduct a validity study to ascertain the psychometric properties of the Spanish translation of the intrinsic and extrinsic motivation (IM and EM) subscales of the Motivated Strategies for Learning Questionnaire with the modified response scales validated in the Danish higher education context. The items
of the IM and EM scales were translated using a forward-backward approach, as where the response scales. To compare the psychometric properties to the only European validity study of the scales using Rasch measurement models, the ordinal Rasch model (RM) and graphical loglinear Rasch models (GLLRM) were employed. To extend on the findings of the previous study, emphasis was on measurement invariance (i.e. no differential item functioning) relative to
courses attended and university. The sample consisted of 875 Psychology students enrolled in the same four subject courses of the Bachelor in Psychology programs at two major Spanish Universities. Data was collected in class one month into the courses. Neither the IM or the EM scale fit the pure RM. The IM subscale fit a GLLRM including DIF relative to course targeted on one item and local dependence between two items. The EM subscale fit a GLLRM including
DIF relative course targeted for one item, DIF relative to gender for another item and local dependence between two items. Targeting and reliability varied for subgroups defined by DIF variables. Reliability was low for most subgroups on both subscales. Targeting was good for the EM subscale, while poor for most
subgroups on the IM subscale.