Alsarmi, Khalifa Salim
(2013)
Psychometric Properties Of The Geography Achievement Test, Geography Self Efficacy Scale, Usm Emotional Quotient Inventory And Their Relationships.
PhD thesis, Universiti Sains Malaysia.
Abstract
To draw accurate conclusions about student achievement and its factors, it is very important to employ instruments with good psychometric properties. This study aimed to examine the psychometric properties of three instruments namely: Geography Self-Efficacy Scale (GSES), USM Emotional Quotient Inventory (USMEQ-i) and Geography Achievement Test (GAT). Additionally, this study aimed to examine the relationship among geography self-efficacy, emotional intelligence and geography achievement of 12th Grade students in Oman. Sample of the study consisted of 839 12th Grade students from two districts selected by cluster sampling from ten schools. Data were collected using three instruments: GSES, USMEQ-i and GAT. Data were analyzed by employing Rasch measurement model to examine dimensionality, fit statistics, gender DIF, response categories, reliability and construct validity of these instruments. Additionally, the structural equation modeling technique was used in order to examine the relationships among the geography self-efficacy, emotional intelligence and geography achievement. Findings showed that all scales in this study met the assumption of unidimensionality. All items in each instrument fit the expectation of Rasch model. The five response categories in GSES and USMEQ-i found to be adequate. Some items in each instrument showed gender DIF. Reliability indexes showed that the item reliability for all instruments and their sub-dimension was high, while the person reliability was good for over all instruments and low for some sub-dimensions such as Emotional Commitment from USMEQ-i and Higher Thinking Level from GAT. Person- item maps of these instruments did not show any significant gap in items distribution of GSES while it showed one significant gap in USMEQ-i and three in GAT. It also indicated that the items of GSES and USMEQ-i were easy and did not match the students’ abilities well. Findings from structural equation modeling demonstrated significant and positive direct relationships between emotional intelligence and geography self-efficacy as well as between geography self-efficacy and geography achievement. The direct relationship between emotional intelligence and geography achievement was found to be not significant. Additionally, the indirect relationship between emotional intelligence and geography achievement through geography self-efficacy was discovered. Based on these results, future research recommendations include reviewing these instruments in order to improve their psychometric properties, validating other instruments related to emotional intelligence and self efficacy in the context of Oman, and replicating this study in Oman with other grade level and different school subjects.
Actions (login required)
 |
View Item |
