الخصائص السيكومترية لمقياس أكسفورد للسعادة باستخدام النظرية الكلاسيكية ونموذج التقدير الجزئي لدى عينة من طلبة جامعة تعز
This research aimed to compare between classical test theory and partial credit model by studying the psychometric properties of happiness Oxford scale for sample of Taiz University. The comparison was focus on: 1) The relationship between the two procedures of the item were selected. 2) To what extent the item statistics (Item difficulty and Item discrimination) are comparable?. 3) To what extent the reliability coefficient derived from the two procedures are comparable? The sample consisted of (906) students of Taiz University for 2017-2018, distribution to four collages (two literary and two scientific) of 508 male and 398 female, the result showed that the existence of high degree of conformity and congruity between the two procedures in estimating the difficulty of vocabulary, the correlation coefficient between the two procedures was high (0.93), as the same time the conformity rate in the selection of items was not high and estimated (0.76). The number of items excluded by Partial Credit Model was (8) while the number of items excluded by Classical Test Theory was (4), in addition to that both of two procedures were congruent in excluding (3) vocabularies. As for stability indicators of the scale, the results indicated that the stability coefficient obtained from the partial credit model was higher than classical test theory, the stability coefficient obtained from the partial credit model estimates (0.98) while it is (0.83) from classical test theory. (Published abstract)