home *** CD-ROM | disk | FTP | other *** search
- There are a lot of statistics presented by !Menstat and many command
- options available. The idea is to allow the user to most effectively
- predict menstrual cycles. The method used is to collect the data, ie. the
- history of an individual woman's cycles. This data is called the sample.
- Then !Menstat calculates statistics to describe the sample. Some
- statistics are intended to estimate the most likely length of a typical
- menstrual cycle. These are called estimators and there are four used
- here:
- mean The average cycle length.
- mode The most frequent cycle length.
- median The middle of all the cycle length sorted in order.
- midrange The average of the longest and shortest cycles.
- Look at the Statistics window to see a chart of the cycle lengths and a
- report of the estimator values. These estimators are used to create the
- Estimates window report of the likely dates of menstruation. From
- 20 to 45 days is considered usual, and 28 days seems to be the most
- typical.
- The only thing a person can actually predict about menstrual cycles is
- that they tend to vary. The standard deviation describes how much the
- sample varies. A standard deviation of 3 to 7 days is typical, and the
- lower the standard deviation, the less that the record cycles vary and
- so the more accurate that future cycles will be to predict.
- Another statistical approach is to plot the recorded dates and then draw
- a straight line through the middle of the plot. This is called linear
- regression. Look at the Plot Cycles window to see the cycles and their
- linear regression plotted on a chart. The linear regression can also be
- used to estimate future cycles by "extending" the line on the plot across
- dates into the future.
- The point is to chose the best estimates by viewing the two charts
- mentioned here and then using the Options menu to select the best
- estimator. The Estimates window and Cycle Chart will update
- automatically as new options are chosen.
- The Options menu provides other capabilities for people who are
- familiar with using statistics. For example, outliers can be filtered
- by using the Constraints option. There are more features, so if you
- like working with statistics then please explore.
-