Table 4.

Minimum sample sizes for adequate statistical power to detect associations of varying effect sizes with latent variables in SEM.

Number of latent variables	Minimum sample size
	|ρ| = .1	|ρ| = .2	|ρ| = .3
1	87	22	10
2	152	37	16
3	290	69	28
4	387	92	38
5	463	110	45
6	526	124	50
7	579	137	55
8	625	148	60
9	666	157	63
10	703	166	67

|ρ| = the smallest effect (Gini correlation coefficient) to detect on latent variables.

Note: Assuming an alpha-significance criterion of α = .05 and power = .80. See Westland (2010, 2012) for the equation for estimating minimum sample size when predicting other variables. The minimum sample size refers to the minimum sample size to detect an association between latent variables. The actual lower bound sample size necessary is the greater of the two values from (a) the minimum sample size necessary for a measurement model (see formula in Westland, 2010) and (b) Table 4 (here).