Significance testing

The output reported by the maximum likelihood does not equal the true parameters or the true volatility, but an estimate thereof.

Which of the following answers is incorrect.

50 XP

The standard error quantifies the estimation uncertainty.
press1
A t-statistic is useful to test whether the estimated parameter is sufficiently different from zero to conclude that the true parameter is different from zero.
press2
The rule of thumb is to conclude for significance when the p-value is less than 5%.
press3
A standard error can take any real value between minus and plus infinity.
press4