3 ms·

That's why your analysis shouldn't report posterior probability but degree of shift in confidence.