This seems to depend almost entirely on what other models you had. A 1% belief in a wider model (say one using a Cauchy distribution rather than a normal one) might have been sufficient to make the result considerably less surprising.
This seems to depend almost entirely on what other models you had. A 1% belief in a wider model (say one using a Cauchy distribution rather than a normal one) might have been sufficient to make the result considerably less surprising.