Open Conference Systems, MISEIC 2020

Font Size: 
Psychometric Properties based on Rasch Modeling of Student’s Meta-affective and Meta-Cognitive in Science Learning
Lilit Rusyati

Last modified: 2020-06-21

Abstract


Every individual has different variety way for correct their self-performance and metacognitive ability is responsible for this. Complex meta-knowing capabilities in young generation come out in the end of a developmental progression so that many adults are not take control in that situation. Then, students’ awareness on emotional engagement gives effect whether positively or negatively during learning and this is reflected by meta-affective. The present study aims to analysis by using Rasch Modeling on meta-affective and meta-cognitive instruments which has been adapted in Bahasa Indonesia version. This analysis focus on item quality and reliabilities for these instruments. An instrument set of Meta-Affective Trait Scale (MATS) and Metacognitive Awareness Inventory (MAI) was distributed to 205 Indonesian secondary students (7th and 8th class). These questionnaires are using Likert scale, MAI with a five-point response scale, where 1=Not at all typical of me, 2 = Not very typical of me, 3 = Somewhat typical of me, 4 = Fairly typical of me, and 5 = Very typical of me while MATS on a 6-point rating scale, from 1 (never) to 6 (always). There are two dimensions of MAI namely knowledge and regulation, while MATS consists of awareness and regulation dimensions. Data were analyzed using Winsteps Rasch Model. The finding describes that there is one misfitting item on dimension “knowledge†of meta-cognitive instruments (MAI) with score 1.51 since range of item fit is 0.5-1.5. Meanwhile for item reliability categorized by “very good†(0.91 - 0.94), and excellent (> .094). To sum up, both MAI and MATS in Bahasa Indonesia version suitable as instrument for assessing meta-affective and meta-cognitive in Indonesian students. Even though there is one misfitting item, but the score in tolerance range. Moreover, supporting by high reliabilities index, so that these instruments fit for other research process.

 


Keywords


Meta-affective; Meta-Cognitive; Psychometrical properties; Rasch Modeling.