. Author manuscript; available in PMC: 2013 Dec 1.

Published in final edited form as: Brain Imaging Behav. 2012 Dec;6(4):517–527. doi: 10.1007/s11682-012-9176-1

Table 2.

Z-statistics for time from mixed models for the EF outcomes, controlling for age, education, gender, and any APOE-ε4 alleles. Sample size needed per group to detect a 25 % decrease in average annual change, with 80 % power and alpha=0.05, two-sided

EF score	Time	N per group^b
ADNI-EF	−9.77	2218
Category Fluency	−9.15	3808
Clock Drawing	−3.42	41,884
WAIS-R Digit Symbol	−6.85	6095
Digit Span Backwards	−4.13	32,162
Trails A^a	−6.59	7798
Trails B^a	−8.17	5082
Trails B–Trails A^a	−6.23	11,187

Trails scores were reversed so higher scores are better for all outcomes

Sample size calculations are based on the coefficient for year and the adjusted residual standard deviation from the full model. As such, they do not directly correspond to the z-statistics