N-SEA Documentation: Systematic Sampling

A very simple sampling technique is to take every n^th entry from a data set to form a Systematic sample (where n is altered according to the size of the sample required). This is a quicker alternative to random sampling when done by hand, but can be biased if the data already contains systematic trends. For example, football attendance figures for a club that plays on Saturday and Wednesday each week may well have alternating high and low attendances.

To obtain a systematic sample then select the 'Statistics' and 'Sampling' menus from the N-SEA menu. When the initial dialogue box appears and the 'Systematic' option is chosen the dialogue box will look like that below.

Click on the 'Data to be sampled box' and then select the data to be sampled from the Excel worksheet. The data should comprise all of the columns of data if you intend to choose the copy options in the second dialogue box. If you intend to construct a filter then this is not necessary.

You should type in a number in the 'Take a sample of' box and define how this is to be interpreted by selecting one of 'percent of observations (rows)', the default, or 'observations (rows)'. If you elect to interpret the number as a percentage then it should be greater than 0 and less than or equal to 100. If the number represents a number of observations (rows) then it must be positive. If it exceeds the number of data items in the complete data set, then the complete data set is returned as the sample.

If you press the 'Next' button then output options are presented in a dialogue box. The 'Finish' button takes the default output options.

This output box and the options have been described in greater detail under 'Random Sampling'.

© Numerical Algorithms Group

Visit NAG on the web at:

www.nag.co.uk (Europe and ROW)
www.nag.com (North America)
www.nag-j.co.jp (Japan)

http://www.nag.com/n-sea/docs/systematicsampling.asp