Investigation of measurement invariance of PISA 2015 collaborative problem solving skills according to countries: Turkey, Norway, Singapore

Purpose of the research is; examine PISA 2015 Cooperative Problem Solving to intercultural measurement invariance of cooperative problem solving model created by Xandar sub-test. within this scope the research was conducted total 2.990 people which was from 1.032 people from Turkey, 923 from Norway and 1.035 from Singapore. Exploratory factor analysis was first performed in order to see the factor distributions of the substances applied in cooperative problem solving and to determine whether they are in the structures specified in the literature. The cooperative problem solving model revealed by exploratory factor analysis was confirmed by confirmatory factor analysis with 7 items and 2 factors. The model data compliance level of the verified model is that it conforms well to the level of significance p = 0.05. At the same time, the values of X2 / df (3,127), RMSEA (0,027), CFI (0,987) and TLI (0,979), which were examined in order to control the effect of sample size, showed that the model was well matched. In the next step, it was investigated whether the model provides measurement invariance between cultures. Invariance analyzes were analyzed by Multi-Group Confirmatory Factor Analysis. The findings show that the cooperative problem solving model provides only structural invariance between cultures and has not provided the metric invariance phase. Therefore, scale and strict invariance stages could not be tested. Obtained results; show that the comparison is not meaningful between the country and test the factor loadings, variances, reveals that vary in the error variance and covariance.