in reply to
Re^6: Random sampling a variable length file.
in thread Random sampling a variable record-length file.
...Just as the polster might discover that the "random selection" they make of the populous, happens to coincidentally consist of the entire membership of some extremist political organisation, they cannot know it will until they take the sample.
A valid sample may contain every member of an extremist group or only contain records whose lengths are even numbers. What makes the sample valid is the plan for selecting the sample (and adherence to that plan). Developing a sampling plan is non-trivial. It can be aided by things we know theoretically and by observational data. But sometimes you just get atypical samples.
In this case, I was confused by your statement in another response that "there is no meaningful correlation in the ordering of the records." This sounds like a positive assertion that there is no correlation based on order. This is quite different from the negative assertion that you don't know whether there are order related correlations.