TABLE 2.

Pairwise comparison of the variability of the estimates of the pRF size between estimation methods at different levels of noise (narrow pRF: angle = 0.59 radians, eccentricity = 1.79°, pRF size σ = 0.23°).

	Low Noise				Medium Noise				High Noise
	GS	COpRF	LinEnc	ModelAve	GS	COpRF	LinEnc	ModelAve	GS	COpRF	LinEnc	ModelAve
HCP7pRF	n.s.	1.24	5.56	9.10	n.s.	1.12	1.67	12.53	n.s.	1.13	n.s.	3.93
GS		1.24	5.61	9.16		1.12	1.68	12.62		1.14	n.s.	3.87
COpRF			4.48	7.33			1.48	11.17			n.s.	3.47
LinEnc				n.s.				7.45				3.84

Every cell displays (for significant effects) the variance of the pRF size estimated by the method in the row divided by the variance of the method in the column ($\frac{\mathit{Var}{(\mathrm{\sigma })}_{\mathit{row}}}{\mathit{Var}{(\mathrm{\sigma })}_{\mathit{col}}}$), while n.s. indicates a non-significant effect.