Stress and prospective memory in the Fourth Age: Evidence fromthe ADuLTS Study

Changes in intelligence over the lifespan is an area developmental researchers have explored for decades. Methods proposed to examine these age-related differences, most notably by Horn and McArdle (1992), examine the relationship between age and factors of intelligence instead of examining the differences in subscale performance. The aim of the current study is to examine the metric invariance across age groups of eight WAIS-R subscales using a two-factor structure. Results suggest that for most of the subscales, metric invariance holds, but on the crystallized intelligence factor, older adults perform worse on the picture arrangement subtest, and on the fluid intelligence factor, older adults perform worse on the block design subtest. In both factors, older adults had the highest variance compared to the other age groups, suggesting that older adults have more variability in crystallized and fluid intelligence than younger age groups. The final model proposed yields both acceptable chi-square fit at the p < .01 level (df = 106; chi-square = 135.69, p = .027) and acceptable fit in most of the alternative fit indices (RMSEA = .017, NNFI = .99, CFI = .99, SRMR = .056). Implications for future research will be discussed.