| Popis: |
The research, by using Trends in International Mathematics and Science Study (TIMSS) 2015 data, was conducted on the data set of 7,223 students who received the booklet number 11. 8th grade math achievement test scores of the countries, test statistics, normality tests and reliability coefficients were examined. It is concluded that the central tendency and distribution measures are close to each other. Confirmatory factor analysis was performed for each country separately and it was concluded that χ2 / sd ratio was quite good. When the fit indexes were evaluated in general, all countries' measurement models were verified separately for all groups. When we take the results of the equality test of covariance matrices for mathematics achievement test scores of The Organısatıon For Economıc Co-Operatıon and Development (OECD) member countries, the fit indices are quite good. Measurement invariance was tested by multi-group confirmatory factor analysis, the values of the fit indices for structural invariance were acceptable and the model was confirmed. The results of the study showed that the measurement invariance could not be achieved for OECD member countries and that the best working model among the measurement invariance models was the strong factorial invariance model. When all the results are evaluated, it is seen that the results of mathematics achievement in TIMSS 2015 are comparable to the results obtained from participants from different countries, languages, genders and cultures and the measurement invariance related to these results cannot be achieved. TIMSS 2015 is open to discussion as the measurement tool does not have the same meaning for all participants. This research draws attention to the fact that measurement invariance cannot be achieved when evaluating the results of large-scale examination applications, and that this situation is taken into consideration carefully when making comparisons and interpretations between countries. |
