Is it possible to have a point estimate outside of confidence interval? Could anyone give me an example?

  • $\begingroup$ Presumably with the qualification that the point estimate and the CI are estimating the same quantity. Otherwise, the answer is obviously yes. $\endgroup$ Apr 4, 2017 at 20:20
  • 1
    $\begingroup$ Just a few hours ago a nice example was posted at stats.stackexchange.com/questions/271845. As the confidence narrows, this particular interval converges to the median of an asymmetric distribution, while the estimate itself is made at the mean of that distribution. Consequently, for a sufficiently small confidence the estimate must lie outside the confidence interval. $\endgroup$
    – whuber
    Apr 4, 2017 at 20:41