ABSTRACT
Background
To be able to conduct comparative research on aphasia assessment and rehabilitation across languages, there is a need for comparable assessment tools. Croatian and Norwegian are among the first languages for which the Comprehensive Aphasia Test (CAT) versions have been published within the Working Group 2 of the Collaboration of Aphasia Trialists network. Comparing the performance of PwA and the psychometric properties of the CAT in these two languages is an interesting test case for the validity of the adaptation process due to the linguistic and cultural differences between the two languages.
Aims
The aim of this article is to investigate whether the goal of creating comparable versions of the CAT – the Croatian (CAT-HR) and the Norwegian (CAT-N) – could be achieved. First, we compare the performance of the two groups of PwA, and second, we investigate the psychometric properties of the two tests and the influence of age and education level on performance. Analogous (psycho)linguistic decisions during test adaptations should secure the comparability of the two versions. Moreover, education level is expected to influence outcome more than age.
Methods and procedure
Two groups of persons with aphasia (PwA), native speakers of Croatian and Norwegian (N Cro = 114; N Nor = 85), were included, as well as two groups of healthy controls (HS: N Cro = 123; N Nor = 84). Data was collected during the norming process of CAT-HR and CAT-N. We compared the performance of HS and PwA on 20 subtests across six language modalities, using nonparametric tests for independent samples. Psychometric properties were assessed through item difficulty and alpha reliability coefficients. The influence of socio-demographic variables on performance was examined using ANOVA.
Outcomes and results
Between-group differences were obtained on six subtests and two modalities, with the Norwegian PwA achieving higher scores on all but one subtest. All HS scored almost at ceiling, but differences were found on twelve subtests; Norwegian HS achieved higher scores on nine. Items in CAT-N have lower overall item difficulty, but alpha reliability coefficients are comparable across subtests. In the Croatian sample, mainly education level influenced performance. In the Norwegian sample, this was observed to a much lesser extent.
Conclusion
Between-group differences can be explained by differences in language structure, reflected also in the item difficulty. On the other hand, reliability of CAT-HR and CAT-N is high and comparable. Implications are discussed in the context of the test adaptation process.
Acknowledgements
The authors thank the Collaboration of Aphasia Trialists, which is funded by COST, and the Tavistock Trust for Aphasia for their support in setting research priorities and promoting international/multidisciplinary collaboration in aphasia research. Furthermore, the authors thank artist Marko Belić for creating new drawings for the different language versions of the CAT, associate professor Bård Uri Jensen for help with the statistics of the Norwegian data, and assistant professor Valentina Ružić for help with the statistics of the Croatian data.
Disclosure statement
The authors report there are no competing interests to declare.
Disclaimers
We hereby declare that the views expressed in the submitted article are based on the research findings and not an official position of the institution or funder.
Notes
1 The difference in cut-off value (115.88 vs 65.75) between the two CAT versions is due to the fact that different modes of computation were employed.
2 The original version of CAT-EN included the DQ as a measure of aphasia impact on the quality of life of the PwA. In the revised, second edition of CAT-EN (Swinburn et al., Citation2023), the DQ is replaced by the AIQ, but no other changes are made..