🚀 go-pugleaf

To drop a mildly humorous item in here...

In 1966 I was a Research Associate in the Department of Epidemiology
at the School of Public Health (Univ. of North Carolina, Chapel Hill)
We had our usual flock of students with their feverishly-keypunched 
card decks containing encoded responses from their surveys and 
questionnaires.

As you may well suspect, there were many errors buried in their
data, so I took it upon myself to create a program which would interpret
the various codes in their punched cards (with only 80 columns, one
tended to get very "cute" when encoding information), check each
data field for valid values, and translate it into a "longer" data
record which could then be fed into the various "canned" statistical 
packages.   This wonderful tool was called (in case you didn't
already see it coming)....

    Epidemiology
    Department
    Interpretive
    Translator

("EDIT" for short)

But that's not the end of the story...

It seems there was a rather large study performed a few years
before I arrived.  More than one student obtained his Ph.D. from
the great insights and wisdom extracted from the results of
this study, and they only had "primitive" card sorters and
tabulating machines to work with.  Then came along another student
whose research topic required that he use the many drawers of
punched cards from this massive study, but he had the advantage of
using a computer to quickly give him many correlations between
various fields.  

When he started getting some strange "blips" outside of the
expected curves, he went digging to find what was causing them.
Among other things, he found data cards giving:

  - a 350-pound man who was 3 feet tall
  - a 12-year-old man who began menstruating at age 26
  - a 90-year-old woman who had given birth to 30 children

You get the idea!  And this was the supporting information for
learned articles long-since published!

So he ran the entire multi-drawer card file through EDIT, went back
to the original paper questionnaire forms, corrected the errors,
and *then* was able to proceed with his own study.

The next day after the story of the "published bad data" made the
rounds of the Department, the following appeared on the door to the
room with all the cabinets of punched cards...

      Never trust another's data.
      EDIT now, or suffa latah!

Cheers,
  Lindsay

Lindsay Cleveland         Digital Systems Co.   Atlanta, Ga
  gatech!dscatl!lindsay     (404) 497-1902
                         (U.S. Mail:  PO Box 1149, Duluth, GA  30136)