The conversion between the effective pRF size (as inferred from the anatomical data) and the Gaussian sigma used by the model is currently very straightforward: Gaussian_sigma = effective_pRF_size * sqrt(output_nonlinearity). However, the actual relationship between them will depend on the other parameters of the model and so we want to test the relationship and come up with a more robust correction.