bootstrap by psignifit

サンプルデータによるbootsrap法の適用例１
>>pfit(dat)
psi(x) = gamma + (1 - gamma - lambda) * F(x, alpha, beta)
using logistic function for F, and assuming 2AFC design (gamma = 0.5)
lambda is constrained within [0, 0.05]
fitting to original data
initial: {alpha = 2.75851, beta = 0.47464, gamma = 0.5, lambda = 0.01}
final: {alpha = 2.78305, beta = 0.482604, gamma = 0.5, lambda = 1.02053e-06}
running 1999 simulations using fitted parameters
Stats for generating distribution:
D = 4.63268, r_pd = 0.148682, r_kd = 0.0515378
random seed: 952975053
2 seconds.

running sensitivity analysis (8 points)
15 seconds

% PARAMETERS:
% alpha beta gamma lambda
2.783 0.4826 0.5 1.021e-06 % (MLE)

% THRESHOLDS:
% 0.2 0.5 0.8 conf
2.1140 2.7830 3.4521 % (MLE)
% BOOTSTRAP THRESHOLD LIMITS:
-0.5973 -0.2997 -0.2913 % 0.023
-0.2767 -0.1396 -0.1336 % 0.159
+0.2114 +0.1419 +0.1941 % 0.841
+0.4185 +0.2585 +0.4110 % 0.977
% WORST-CASE THRESHOLD LIMITS:
-0.6944 -0.3078 -0.3261 % 0.023
-0.3423 -0.1603 -0.1543 % 0.159
+0.2422 +0.1756 +0.2691 % 0.841
+0.5012 +0.3245 +0.5809 % 0.977
以下省略