statistical thinking

Six Sigma – iSixSigma Forums Old Forums General statistical thinking

Viewing 3 posts - 1 through 3 (of 3 total)
  • Author
  • #50497


    can any body explain me why we need to do ststistical tests on data with an example…



    Is your question on this site normal variation (standard process) or an outlier (special cause)? Count the number of times your question has been asked in the last year. Count the number of posts for this year and get the ratio. Trend this info and see whether I should have reacted or not…. how many times a day are you effected by reactive changes (thing about the gas pump).



       Data with an example is much more difficult to deal with than data without an example.  Data that is free of examples is data that is clean.  Clean data is certain data, therefore any conclusions one might draw from such data are absolute and without error.  However, if there are one or more examples in the data then the only way you can hope to use this kind of data as a basis for any conclusion is if you perform one or more ststistical tests on the data to remove or at least minimize the effect of the example.  There are many ststistical tests one can use with example laden data – TARFU, FUBAR, FIGMO, and SWAG procedures are probably the best known tools but the best method for analysis of this kind of data is either generalized BWBS or the newer DATAFOG algorithms.

Viewing 3 posts - 1 through 3 (of 3 total)

The forum ‘General’ is closed to new topics and replies.