As a budding researcher so many years ago I discovered quite by accident
how easy it is to draw the wrong conclusion by injecting my own bias into
a study.
I was performing an undergraduate genetic study using fruit flies. These
flies had "known" mutations, but I did not know exactly what mutations my
flies had. To find out, we were required to breed them for many
generations and make observation to determine what the mutations were.
Well, after my F1 or even my F2 generations, it was clear to me that the
ratios in my data had revealed what my mutations were, and I, thinking I
had this genetic stuff down, was quick to announce I had it all figured
out. I then thought, that since it wasn't taking very long to produce
each generation, I would just run a few more generations to confirm my
new discovery.
You see where this is going don't you!!! With each succeeding generation,
the data were not confirming my "suspected" ratios! Panic set in now. I
kept producing generation after generation, and the data would just not
confirm my original conclusion! I was now over 1000 generations worth of
data and truly stumped, so I approached my instructor to ask for advice.
His advise was, "Ed, you have plenty of data, and you have tried your
damndest to tell the data to confirm your hypothesis, now it is time to
let the data tell YOU what is really there!"
Well, low and behold, my original "expectation" was far to simplistic,
thinking there was only one mutation in these flies, when in fact there
were actually multiples.
What I got out of this was just how easy it is to let ones bias affect
ones findings, the importance of sample size as it pertains to accuracy
of ones findings, and accurate data collection ( which I find can be the
most tedious job, but is imperative if one is to expect any real value in
their findings). If I had stopped my experiment after just the first
couple observations, I would have been DEAD WRONG! So please don't be
quick to make a judgement after seeing something once or twice.
Without the scientific method applied in a rigorous fashion, what we end
up with are the "Old Wives Tales" of beekeeping.
____________________________________________________________
Stock Options
Click to learn about options trading and get the latest information.
http://thirdpartyoffers.juno.com/TGL2141/c?cp=5juI2U6-rckNm5MrkFR0xAAAJ1CYh4LMlrx5Sko2bwEETtn7AAYAAAAAAAAAAAAAAAAAAADNAAAAAAAAAAAAAAAAAAAQlgAAAAA=
***********************************************
The BEE-L mailing list is powered by L-Soft's renowned
LISTSERV(R) list management software. For more information, go to:
http://www.lsoft.com/LISTSERV-powered.html
Access BEE-L directly at:
http://community.lsoft.com/scripts/wa-LSOFTDONATIONS.exe?A0=BEE-L
|
