Periodic Sample command creates a systematic sample for each input variable: it extracts values from the input variables, starting at the specified index and then extracts every i^th value, where i is the sampling interval or the skip. Periodic sampling is also known as a systematic sampling. If an input variable has periodic patterns, the sample will be biased.

How To

Run: Data→ [Data Sampling] Periodic Sample...

Select variables. The first random sample is drawn from the first variable, then the second random sample is drawn from the second variable, and so on.

Specify the interval of sampling (skip): enter a number to the Interval/Skip field. The interval of sampling can be found by dividing the input variable size (number of observations) by the desired sample size.

Enter the index of the observation with which to start the sampling: enter a number to the First Observation Index field.

Results

The table with sample values is generated. The number of values in the table column is equal to the number of values in the input variable, divided by the sampling interval.

 

References

Lohr, S. (2010). Sampling: Design and analysis, 2nd ed. Boston: Brooks/Cole.