. Author manuscript; available in PMC: 2010 May 26.

Published in final edited form as: Neuropsychology. 2008 Jul;22(4):531–544. doi: 10.1037/0894-4105.22.4.531

Table 2.

Relative Fit Indices for Each of the Models Tested

Model	χ² (DF)	CFI (>.95)	TLI (>.95)	RMSEA (<.O8)
1 Factor Model (glob)	1273.1 ( 69)	.931	.989	.172
3 Factor Model (glob, mem, nonmem)	891.5 (105)	.955	.995	.113
5 Factor Model (glob, mem, lang/sem, vsp, exec)	606.5 (133)	.973	.998	.078
6 Factor Model (glob, mem, lang, sem, vsp, exec)	596.7 (133)	.973	.998	.077
7 Factor Model (glob, mem, lang/sem, vsp, plan, org, div att)	460.2 (139)	.982	.999	.063
8 Factor Model (glob, mem, lang, sem, vsp, plan, org, div att)	452.1 (139)	.982	.999	.062

Notes. glob = global; mem = memory; nonmem = non-memory; lang = language; sem = semantic knowledge; vsp = visual spatial; exec = executive; plan = planning; org = organization; div att = divided attention; CFI = comparative fit index; RMSEA = root mean square error of approximation; TLI = Tucker-Lewis Index. The TLI and CFI range from 0 (poor fit) to 1 (perfect fit); values of .95 or higher are indicative of a good model fit. RMSEA values lower than .08 are considered to reflect adequate fit, values less than .05 to .06 indicate good fit.