Computerized lab notebook?

Craig V Lewis cvl at athena.mit.edu
Sun Jan 10 14:39:26 EST 1993


I strongly recommend against the use of a computer as your sole data record!!

They are far too sensitive to several hazards; magnetic fields, physical
shock, liquid spills, and software viruses may all eliminate data
irretrievably.  There exist commercial programs capable of wiping data
from hard disks so completely that they meet government security standards.
Just imagine spilling a critical sample into your computer immediately
after analysis.  The sample is gone and the data are gone;  you will have to 
repeat the experiment, wasting valuable time.  Now try the same trick on
your paper lab notebook; any reasonably well chosen ink will not run 
significantly and you spend 5 minutes with a hairdryer to resurrect your
data.

In addition, most commercial computers do not provide a secure time stamp
for the data.  A skilled computer operator can very quickly modify creation
dates and data without leaving any trace of doing so.  Paper notebooks are
much harder to modify without leaving at least some suspicious traces, and 
the art of analyzing handwritten and typed documents has been very highly
developed.  Interested parties should read the David Baltimore case for 
examples of the analysis techniques applied to lab notebooks.

I hope I've made the case strongly enough, because I feel it is of critical
importance.  I work on computers 8-12 hours per day, but I never trust them
with anything utterly essential without having a printed back-up or stored
copies in a wide variety of locations.

Which reminds me, I have to print some things out now. : )
--
 Craig Lewis                            cvl at athena.mit.edu                   
 Woods Hole Oceanographic Institution   (508)457-2000x4814                   
 Nobody but me has any idea what I'm doing; and I don't know either.         



More information about the Comp-bio mailing list