Table 1.

Summary of the population averages of measures of contrast-dependent clustering

Type of data set and unitsNo. of unitsNo. of units: Hmax > HcountΔH/HcountHcountHmax
Dspike,circFallFevenFodd
Edge
 All22150.470.38  ± 0.260.56  ± 0.230.56  ± 0.250.55  ± 0.270.48  ± 0.22
 S1160.330.36  ± 0.270.48  ± 0.190.45  ± 0.210.45  ± 0.210.45  ± 0.22
 Cx1190.600.40  ± 0.270.64  ± 0.240.66  ± 0.260.66  ± 0.290.50  ± 0.23
Grating
 All2290.070.57  ± 0.270.60  ± 0.260.60  ± 0.260.59  ± 0.270.60  ± 0.26
 S1150.080.53  ± 0.270.58  ± 0.240.57  ± 0.230.55  ± 0.250.58  ± 0.23
 Cx1140.050.60  ± 0.280.63  ± 0.300.62  ± 0.300.62  ± 0.290.62  ± 0.30
  • ΔH = Hmax − Hcount, the contribution by temporal structure, obtained with the metrics Dspike,circ[q], is averaged across cells before taking the ratio (ΔH/Hcount). Hmax, the average maximal level of clustering, is listed for the metrics Dspike,circ[q] and three families of metrics based on cumulative use of Fourier components: Fall, Feven, and Fodd. The level of chance clustering is subtracted in all cases.