Longitudinal Invariance of the Wechsler Intelligence Scale for Children–Fourth Edition in a Referral Sample

Abstract

Measurement invariance of the Wechsler Intelligence Scale for Children–Fourth Edition (WISC-IV) was investigated with a group of 352 students eligible for psychoeducational evaluations tested, on average, 2.8 years apart. Configural, metric, and scalar invariance were found. However, the error variance of the Coding subtest was not constant across time, allowing only partial strict invariance. This indicates that the WISC-IV (a) was measuring similar constructs at both test occasions, (b) constructs had the same meaning across time, (c) scores that changed across time can be attributed to change in the constructs being measured and not to changes in the structure of the test itself, and (d) measures the same constructs equally well across time with the possible exception of Processing Speed due to the noninvariance of the Coding subtest’s residual variance. This investigation provided support for intelligence as an enduring trait and for the validity of the WISC-IV.