Cross-Cultural Validation of the Test of Everyday Test of Attention With Confirmatory Factor Analysis

Abstract

The Test of Everyday Attention (TEA) was developed in improve upon existing methods of assessing attentional problems in clinical practice. It consists of 8 subtests measuring sustained, selective, and divided attention. Its construction was also designed to mimic everyday activities. However, the construct validity of the TEA has not been fully studied with a more stringent methodology. This study aimed to examine the construct validity and its stability across cultures with confirmatory factor analysis (CFA). A sample of 148 healthy UK and 133 Hong Kong Chinese participants were recruited. They were well-matched with gender proportion and educational level. Three models were tested in the present study: (1) the one from the original proposed underlying attentional contructs (Visual Selection, Sustained Attention, and Switching) (Robertson et al., 1996); (2) the discovered 4-factor structure by exploratory factor analysis (Robertson et al., 1996); and (3) the discovered 4-factor structure by exploratory factor analysis (Chan et al., 1999). The CFA solutions suggested that the 3-factor model provided the best fit for both samples (AGFI = 0.83, GFI = 0.91, CFI = 0.91 for UK sample; AGFI = 0.85, GFI = 0.93, and CFI = 0.90 for HK sample). A direct comparison of chi-squares further indicate significant differences among the 3-factor models and the alternative models (x2 = 6.54, p = 0.075 for UK sample, and x2 = 25.3, p = 0.001 for HK sample). Therefore, the 3-factor model provides the best fit of attentional components embedded in the TEA and it is consistent and stable across cultures.