I fear my data sample is too small to answer that, yet I have discovered a most interesting point.

Note: Percentages may not add up to 100 percent due to rounding of data samples.

Note that the two-sample test checks whether the two data samples come from the same distribution.

For a finite data sample, the mode is one (or more) of the values in the sample.

These results were confirmed when the analysis was repeated on a larger data sample.

Such intelligence was coming in at the rate of 20 million data samples per second, 24 hours a day.

In such cases, for a safe operation, the system has to be compensated for missing data samples.

Like everything else, it's all a matter of interpreting a mighty slim data sample, Skip.

The characteristics of the data sample can be assessed by looking at:

Also, one observation for New Zealand in 1951 moved the data sample's central tendency significantly away from what it otherwise would have been.