On 8/12/2010 2:49 PM, Andrew Sullivan wrote:
The basic problem is that the survey sample is
self-selected.
That's one of several methodological problems. It also is distributed to a
distinctive group, probably much narrower and less representative of the
population of interest than is intended.
For an established activity, a strategic point of weakness is to use the
existing base for research. While current, dedicated "customers" are always
important, relying only on information from that group means that differences
from a wider, desired audience will be missed.
There were also some significant limitations with the questions that were asked.
Good survey methodology is extremely difficult.
One of the greatest dangers to a bad survey is that it produces good-looking
results that lull folks into thinking they've got useful data.
d/
--
Dave Crocker
Brandenburg InternetWorking
bbiw.net
_______________________________________________
Ietf mailing list
Ietf@xxxxxxxx
https://www.ietf.org/mailman/listinfo/ietf