Anita PásztorKovács, Attila Pásztor & Gyöngyvér Molnár: The Assessment of Teamwork Skills: Validating the Collaborative Skills Questionnaire Anita PásztorKovács, Attila Pásztor & Gyöngyvér Molnár In this paper, we report about our study in which we developed a new assessment tool called Collaborative Skills Questionnaire (ColSQ). We aimed to create an instrument which (1) can be considered as an effective tool for assessing collaborative skills of upper primary school students in terms of reliability and dimensionality; (2) can serve as a basis of convergent validity studies connected to CPS instruments; (3) is appropriate to deepen our knowledge about the structure of CPS competence. as it is based on one of the CPS theoretical models. 2128 Grade 8 students (M=14.7 years; SD=.47) participated in the online data collection via the eDia diagnostic platform. ColSQ was based on the theoretical model of CPS by Hesse et al. In their framework, they outline three core skills and nine subskills which overall compose the social component of CPS. Four items were assigned to each subskill, so the questionnaire contained 36 items in total. Participants had to rate on a 7point scale how well the 36 statements described them. To test the theoretical model, we implemented confirmatory factor analysis. After excluding 18 items with low factor loadings, the remaining 18 items showed an acceptable fit for a regraded one and a threedimensional model (onedimensional model: χ2=1875.91 df=135; p<.01; CFI=.942; TLI=.934; RMSEA=.078; threedimensional model: χ2=1802.83; df=132; p<.01; CFI=.944; TLI=.935; RMSEA=.077). The threedimensional model had a significantly better fit than the onedimensional one (χ2=87.178; df=3; p<.01). The reliability index of the 18item scale was high, Cronbachα=.90. Moreover, the 6item participation subscale (.82), the 4item perspective taking subscale (.70) and the 8item social regulation subscale (.76) also had sufficient reliability. Each of the nine subskills were connected to at least one item of the new, shortened ColSQ. Our study resulted in an 18itemscale which, both from the perspective of reliability and dimensionality, proved to be an effective tool for measuring collaborative skills; furthermore, it seems to be a sufficient method in future validity studies relying on CPS assessment tools, as it is a selfreport, computerbased instrument, which investigates the collaborative component on the individual level. Our results can be interpreted as a confirmation of the theoretical model of Hesse et al. regarding social skills. MAGYAR PEDAGÓGIA 120. Number 3. 269296. (2020) Levelezési cím / Address for correspondence: PásztorKovács Anita, Pásztor Attila és Molnár Gyöngyvér, Szegedi Tudományegyetem Neveléstudományi Intézet. H–6791 Szeged, Petőfi Sándor sgt. 32–34. 
