Table 1

Statistical table of results

DatasetMetricConditionsPaired median
difference
MAD95% CIp-value
Figure 6
 aBinary noise (N = 40) IBernoulliRH–ISI0.002 bits/spike0.003[0.000, 0.003]0.0092
 bBinary noise (N = 40) IBernoulliCH–RH0.009 bits/spike0.008[0.004, 0.015]0.0002
 cBinary noise (N = 40) IBernoulliCH–ISI0.004 bits/spike0.004[0.003, 0.009]0.0002
Figure 7
 dGratings (N = 33) IBernoulliRH–ISI0.030 bits/spike0.020[0.012, 0.047]0.0002
 eGratings (N = 33) IBernoulliCH–ISI0.049 bits/spike0.033[0.032, 0.080]0.0002
 fGratings (N = 33) IBernoulliCH–RH0.020 bits/spike0.014[0.006, 0.027]0.0002
Figure 8
 gGratings (N = 33) IBernoulliQ4–Q1−0.005 bits/spike0.041[−0.047, 0.003]0.353
 hBinary noise (N = 39) IBernoulliQ4–Q10.001 bits/spike0.007[−0.001, 0.004]0.396
Figure 9
 iAnesthetized (N = 27)Absolute differenceGratings–noise 100 ms0.0310.016[0.018, 0.039]0.0002
 jAnesthetized (N = 27)Absolute differenceGratings–noise 30 ms0.0080.007[0.002, 0.012]0.0004
  • CIs are derived from 5000 bootstrap resamples and are bias corrected and accelerated; p-values are derived from paired-permutation tests with 5000 permutations. For details, see Materials and Methods.