VALIDATION OF THE ARABIC-TUNISIAN VERSION OF THE FRENCH COMPETITIVE STATE ANXIETY INVENTORY-2 REVISED (CSAI-2R): INSIGHTS FROM AN EXPLORATORY FACTOR ANALYSIS

SOFIENE MNEDLA^1,*, NICOLA LUIGI BRAGAZZI^2,3, CARLO CHIORRI⁴, ALI ELLOUMI⁵, WALID BRIKI⁶

¹Higher Institute of Sport and Physical Education of Ksar Said, Department of Education sciences, Mannouba University, Tunisia - ²School of Public Health, Department of Health Sciences (DISSAL), University of Genoa, Genoa 16132, Italy - ³Section of Psychiatry, Department of Neurosciences (DINOGMI), IRCCS San Martino-IST, University of Genoa, Largo Rosanna Benzi 10, 16132, Genoa, Italy - ⁴Department of Educational Sciences, University of Genoa, Genoa, Italy - ⁵Laboratoire techniques et enjeux du corps, Université Paris Descartes, France - ⁶Sport Science Program, College of Arts and Sciences, Qatar University, Doha, Qatar

Aim: The Competitive State Anxiety Inventory-2 Revised (CSAI-2R) is one of the most frequently used instruments when quan- titatively investigating competitive state anxiety in the field of sport psychology research. The purpose of the present study was to validate the Arabic-Tunisian version of the French CSAI-2R questionnaire and to evaluate its validity and reliability.

Materials and methods: Out of an initial list of 490 subjects, participants of the present validation study were 304 Tunisian football players (81 males [26.6%]; mean age 21.58 years (SD = 2.40 years, range 16-31) of different levels (international level: 34 [11.2%]; national level: 238 [78.3%]; regional level: 32 [10.5%]) that volunteered to participate in the study. Data were collected and analyzed in terms of reliability and validity using correlations, reliability, parallel analysis and exploratory factor analysis.

Results: The 2-factor solution accounted for 48%, 47% and 56% of variance for the Direction, Frequency, and Intensity sca- les, respectively. The 3-factor solution accounted for 60%, 57% and 64% of variance for the Direction, Frequency, and Intensity sca- les, respectively. Although the 3-factor solution explained an additional, substantial proportion of variance with respect to the 2-fac- tor solution, the factor structure did not replicate well across scales. Conversely, in 2-factor solution, if items 1, 2, 15, and 16 were excluded, the pattern of loadings was exactly the same for all scales. Hence, we concluded that the 2-factor solution should be retai- ned, as it provided a measurement model for the Arabic-Tunisian CSAI-2R items that was constant across scales.

Conclusion: The Arabic-Tunisian version of the French CSAI-2R in its original version has failed to demonstrate robust and sound psychometric properties. Its reduced version represents a first attempt that can be henceforth used in the Arab world, even if further investigations should enable to improve its psychometric properties.