Development Of a self-regulation scale based on the graded response model (GRM)

Abstract

The study aims at developing a self-regulation scale based on the Graded Response Model (GRM), to achieve the objectives of the study the Self – Regulation Scale (Gaumer Erickson & Noonan, 2016), (22 items) translated into arabic language. The translated version was conducted on a sample of 258 students (123 male, 135 Female) of Al-Ahliyya Amman University Jordan. The construct validity of the scale was examined using the data collected from a pilot sample (30 students), and the whole sample (258 students); where the correlation coefficients between the item and the scale ranged (0.023 - 0.784) and between (0.293 - 0.604) for the pilot sample and the whole sample respectively. The Internal consistency of the scale was examined, for the pilot sample and the main sample revealed the value of Cronbach alpha ranged between 0.890 and 0.826, respectively, and a marginal reliability of 0.87.The data of the study supported the (IRT) assumptions (unidimensionality-, local independence-, speediness-, item test information function, item characteristic curve). On the other hand, concerning the threshold intervals, the mean and standard deviation values confirmed the accuracy of the estimated parameter item difficulty. Four items were omitted from the scale for not matching the GRM, while none of the items showed a differential functioning according to the GPA or the academic level of the participants. The construct validity for final form of the scale was examined where the correlation coefficients between the item and the scale ranged (0.327 - 0.613), and the value of the reliability was.