TIMSS 2019 sekizinci sınıf matematiğe yönelik güven ölçeğinin hizalama metoduna göre ölçme değişmezliğinin incelenmesi

YAKUP DOKUR

TIMSS 2019 sekizinci sınıf matematiğe yönelik güven ölçeğinin hizalama metoduna göre ölçme değişmezliğinin incelenmesi

Thesis Type: Postgraduate

Institution Of The Thesis: Gazi University, Eğitim Bilimleri Enstitüsü, Eğitim Bilimleri Enstitüsü, Turkey

Approval Date: 2023

Thesis Language: Turkish

Student: YAKUP DOKUR

Supervisor: Mehtap Çakan

Open Archive Collection: AVESIS Open Access Collection

Abstract:

In this study, confidence scale scores for mathematics, which is one of the affective features in the TIMSS application, were investigated. In the research, the data of the confidence scale in mathematics from the student questionnaires answered by the eighth grade students who participated in the TIMSS 2019 exam were discussed. In this direction, the aim of the research is to examine the measurement invariance of the eighth grade students' mathematics confidence scale scores who participated in the TIMSS 2019 exam according to the determined countries. This research, which is in the relational screening model, was conducted on seven countries (Chile, Hong Kong, Israel, Qatar, Russia, Turkey and England) in order to represent different cultures. Measurement invariance, which is important in comparing countries with each other, is a validity study. In the study, measurement invariance was tested using multiple group confirmatory factor analysis (MGCFA) and alignment method. When the findings of the multi-group confirmatory factor analysis were examined, it was seen that the confidence scale for mathematics provided structural, metric and partial scalar invariance. It was found that the items "I am generally good at math", "Math is more difficult for me than many of my classmates" and "I am good at solving difficult math problems" provide partial scalar invariance. In the findings of the alignment method, the groups that noninvariance for the model were 39.68% for factor loadings; for intercepts, it was found 46.03%. The mean of the groups that noninvariance was found to be 42.86%. Since this value is above the recommended limits, the measurement invariance of the confidence scale for mathematics in the alignment method did not reach the desired level. In the alignment method, parameters can be interpreted even if measurement invariance is not ensured, as parameters are not like invariance indexes. When all criteria for the confidence scale in mathematics are evaluated together in the alignment method, the item "Mathematics is not one of my strengths" is the item with the highest level of invariance according to factor loadings. According to the intercepts, the items "Math is more difficult for me than many of my classmates" and "I am good at solving difficult mathematical problems" are the items with the highest level of invariance. According to both factor loadings and intercepts, the item "My teacher tells me I am good at mathematics" expresses the item with the lowest level of invariance. As a result of the research, it was found that the said scale, which aims to measure the confidence of eighth grade students in mathematics, provides structural and metric invariance between seven countries with MGCFA; It was found that the scale provided partial scalar invariance for three items. According to the alignment method, it was concluded that measurement invariance was not achieved. In other words, it can be said that the confidence scale for mathematics is not equivalent for seven countries. Therefore, this situation should be taken into account when comparing confidence scores for mathematics for these countries.