Skip to main content
. Author manuscript; available in PMC: 2014 May 1.
Published in final edited form as: Br J Math Stat Psychol. 2012 May 29;66(2):245–276. doi: 10.1111/j.2044-8317.2012.02050.x

Table 4.

Simulation Results: Bifactor model, Null Condition

K N reps M2
RMSEA
rejection rates
df Mean Var Mean (90% CI) .150 .100 .050 .010
test statistic based on full one- and two-way marginal tables (M2)
2 750 935 75 74.4 147.1 .006 (.000,.019) .144 .106 .049 .006
1500 922 75 74.7 147.8 .004 (.000,.013) .153 .104 .044 .007
3 750 995 390 390.1 754.1 .004 (.000,.012) .153 .098 .047 .009
1500 998 390 389.8 706.4 .003 (.000,.009) .140 .093 .039 .007
4 750 998 915 888.7 1838.3 .002 (.000,.008) .057 .030 .009 .002
1500 1000 915 891.2 1703.6 .001 (.000,.006) .059 .035 .010 .003
5 750 989 1650 1601.5 3149.9 .001 (.000,.006) .029 .015 .008 .001
1500 989 1650 1604.8 3318.4 .001 (.000,.004) .033 .014 .007 .001

test statistic based on reduced one- and two-way marginal tables ( M2)
3 750 995 60 59.4 133.5 .006 (.000,.021) .140 .096 .052 .014
1500 998 60 59.4 116.5 .004 (.000,.014) .128 .091 .042 .009
4 750 998 45 45.2 96.6 .007 (.000,.023) .154 .106 .059 .017
1500 1000 45 45.1 86.3 .005 (.000,.015) .157 .097 .042 .007
5 750 989 30 29.8 62.3 .006 (.000,.024) .140 .092 .055 .017
1500 989 30 30.1 60.2 .004 (.000,.017) .151 .095 .047 .013

Notes. K is the number of categories. N is the sample size. The number of valid (converged and stable) solutions are reported under the column heading “reps.” The reference distribution of M2 and M2 statistics is central chi-square with df degrees of freedom. Note that M2 is equivalent to M2 when K is 2.