Bad data (from Mary)

[from Mary Powers 960905]

Bruce Gregory (960904.1150 EDT)

     One person's good data is another person's bad data. If you
     had wanted to demonstrate that subjects who cannot control
     to begin with, do not gain the ability to control as a
     result of sleep deprivation, the data you label bad would be
     labelled good.

But that is another experiment entirely, not a study of control
in practiced subjects during sleep deprivation and the
administration of drugs to counter it. My point was that since
the subjects were not practiced sufficiently, the conditions of
the study were not met, and therefore the data was bad - for that
study. Sure, someone could take that data and perhaps dream up a
different hypothesis to test, for which the data is ok. I
imagine that salvage operations like this go on all the time, but
they do not alter the fact that the data was bad for the original
purpose it was taken.

Mary P.

[From Bruce Gregory (960905.1440 EDT)]

(Mary Powers 960905)

My point was that since
the subjects were not practiced sufficiently, the conditions of
the study were not met, and therefore the data was bad - for that
study. Sure, someone could take that data and perhaps dream up a
different hypothesis to test, for which the data is ok. I
imagine that salvage operations like this go on all the time, but
they do not alter the fact that the data was bad for the original
purpose it was taken.

You'll get no arguments from me! I just wanted to make the
point that data is only good or bad from the point of view of
the purpose for which it was gathered.

Best wishes,

Bruce