Table 3.

Bias for η and ζ predicted as a multiple regression from simulation conditions. N is the number of rows in the matrix in hundreds (i.e., 2 or 5). Embedded is coded 0 for the burst sample matrix and 1 for the embedded matrix.

η	Value	SE	t	p
Intercept	0.00560	0.00065	8.562	< 0.001
N (in hundreds)	−0.00006	0.00016	−0.402	0.704
Embedded	−0.00083	0.00048	−1.737	0.143
Multiple R-Squared: 0.389, Adjusted R-squared: 0.144

ζ	Value	SE	t	p
Intercept	0.00057	0.00233	0.244	0.817
N (in hundreds)	0.00004	0.00057	0.077	0.941
Embedded	−0.00066	0.00171	−0.388	0.714
multiple R-Squared: 0.030, Adjusted R-squared: −0.358