Table 2.

Results of brain-behavior analyses using cluster size and peak t score

Region	R2		Age		Accuracy		RT
	R2	P	β ± SD	P	β ± SD	P	β ± SD	P
Cluster size
L PFC	0.27	0.101	0.60  ± 0.25	0.029	−0.07  ± 0.24	0.786	−0.36  ± 0.22	0.118
R PFC	0.56	0.001	0.76  ± 0.20	0.001	0.17  ± 0.19	0.540	−0.30  ± 0.17	0.091
L PPC	0.72	0.000	0.84  ± 0.16	0.000	0.11  ± 0.15	0.469	−0.18  ± 0.14	0.192
R PPC	0.69	0.000	0.77  ± 0.17	0.000	0.16  ± 0.16	0.328	−0.13  ± 0.14	0.379
Peak t score
L PFC	0.29	0.082	0.44  ± 0.25	0.091	0.11  ± 0.24	0.645	−0.04  ± 0.22	0.852
R PFC	0.53	0.002	0.75  ± 0.20	0.001	−0.21  ± 0.19	0.278	−0.15  ± 0.17	0.374
L PPC	0.56	0.001	0.68  ± 0.20	0.003	0.03  ± 0.19	0.885	−0.10  ± 0.17	0.574
R PPC	0.50	0.004	0.54  ± 0.21	0.019	0.18  ± 0.20	0.391	−0.10  ± 0.18	0.603

Multiple linear regression analyses was used to examine the relative contributions of age and task performance (accuracy and RT) to WM-related activation in the PFC and the PPC. Results of two separate analyses using two different measures: cluster size and peak t score in each activation cluster identified in Table 1. Analysis model R², and corresponding P value, along with β weights, and their corresponding P values, are shown for each region. The analysis shows that in each of these regions, age is a significant predictor of brain activation, even after accounting for changes in task performance with age.