Table 2.

Average estimated information transfer as a function of samples per matrix category (N∕m) for m×m matrices of size m=2,…,10. Matrices filled with data sampled from a uniform distribution using the strict sampling constraint described in the text. The true value for information transfer should be 0%.

$I\widehat{T}$ (%)N∕m	Matrix size (m×m)
	2×2	3×3	4×4	5×5	6×6	7×7	8×8	9×9	10×10
10	4.0	7.0	9.6	11.9	14.1	15.9	17.7	19.3	20.8
20	1.9	3.2	4.4	5.5	6.5	7.5	8.4	9.4	10.2
30	1.2	2.1	2.8	3.5	4.2	4.8	5.4	6.0	6.5
40	0.9	1.6	2.1	2.6	3.1	3.5	3.9	4.4	4.8
50	0.7	1.2	1.6	2.0	2.4	2.8	3.1	3.4	3.7
100	0.4	0.6	0.8	1.0	1.2	1.3	1.5	1.7	1.8