anxiety in Swedish and Finnish grade 3 pupils: Measured by the CTAS

The education systems in Sweden and Finland have different formal and informal testing traditions. A recognised possible adverse effect of testing is test anxiety among pupils and students which may have a negative impact on examination performance. Research into which factors of testing practice affect the levels of test anxiety in younger pupils in real classroom settings is a neglected area internationally yet holds great importance for school practitioners. A cross-national study was conducted to determine whether there are any differences in test anxiety between groups of young pupils in Sweden and Finland, as measured by the Children’s Test Anxiety Scale (the CTAS), and whether these differences are ‘real ’ differences or a result of differential item functioning. The di-mensionality of the CTAS construct is further examined. Exploratory Structural Equation Modelling was used to analyse the data obtained. Partial measurement invariance with respect to nationality and gender was achieved, demonstrating that the CTAS accurately measures latent constructs such as thoughts, autonomic reactions and off-task behaviours in boys and girls, and Swedish and Finnish pupils. No differences were found in the levels of test anxiety experienced by Swedish and Finnish pupils. Girls reported higher levels of autonomic reactions related to test anxiety, but no gender differences in thoughts and off-task behaviours were identified. Methodological limitations and the future implications of the results obtained are discussed