Factor analysis of the Wechsler Abbreviated Scale of Intelligence: Comparison of Lithuanian and U.S. standardization samples and different age groups

Abstract

While evaluating individuals’ cognitive functioning, it is important not only to estimate the general level of their intellectual abilities, but also to describe the strengths and weaknesses in the cognitive functioning exhibited by an individual. Recently, the Wechsler Abbreviated Scale of Intelligence (WASI) has been standardized in Lithuania. This test provides three estimates of intellectual functioning: Full Scale IQ (FSIQ), Verbal IQ (VIQ), and Performance IQ (PIQ). It is important to examine whether this short test, consisting of only four subtests, can provide a reliable and valid information about two specific constructs: verbal and perceptual intellectual abilities. Objectives: The research was designed to determine whether the WASI measures the same constructs across cultures and age bands. Results: Exploratory factor analysis was conducted on the WASI Lithuanian standardization total sample (N = 539; age 6 to 89 years) and six age bands (6–11, 12–16, 17–24, 25–44, 45–64, and 65–89). For each sample, the four WASI subtests were subjected to a principal-axis factor analysis followed by promax rotations. Two factors were specified to be retained. The Verbal Comprehension factor, consisting of Vocabulary and the Similarities, and the Perceptual Organization factor, consisting of Block Design and Matrix Reasoning, were identified in all analyses. The coefficients of congruence were 0.998 for Factor I and 0.993 for Factor II, suggesting factorial equivalence across the Lithuanian and the U.S. standardization samples. The high coefficients of congruence, ranging from 0.920 to 1.000, suggest the stability of factor structure across the age groups. A confirmatory analysis was performed for each of the six age bands. For each of these confirmatory analyses, a two-factor model was compared with the general one-factor model. The results support the premise that the two-factor model best fits the data for 12–89 age bands, whereas the one-factor model best fits the data for 6–11 age band. Conclusions: The structure of the Lithuanian version of the WASI is the same as in the U.S. standardization sample. The structure is stabile in adult samples across cultures. While evaluating individuals aged 12 to 89 years, it is better to interpret two factors (verbal and nonverbal abilities), whereas while evaluating individuals aged 6 to 11 years, it is better to interpret one factor general intellectual ability.