Results based on subgroups have a larger potential sampling error.
Its sampling error could be plus or minus three points.
The potential sampling error for the polls ranged from two to four points.
The term can also be used to mean sampling error in general.
For likely voters, the potential sampling error is five percentage points.
The poll has a possible sampling error of 3.5 percentage points.
Given the sampling errors, these results can be considered identical.
The gap - only half a point - is well within the range of normal sampling error.
The Fed survey, and especially the results for the very rich, is subject to significant sampling error.
The potential sampling error for groups with fewer cases is larger.