***** To join INSNA, visit http://www.insna.org *****
Has anyone done any work comparing the effects of saturation sampling versus
name generator methods on social network measures? More specifically, is
anyone familiar with reesearch that makes comparisons between groups for
which data are collected using different sampling methods?
We are conducting social network research on voluntary civic society
associations. We have begun by studying a sample of neighborhood councils.
Because they are relatively formalized with an elected board of directors,
we have been able to employ saturation sampling. We provided each
respondant with a complete list of board members and asked them to identity
the members with whom they were in contact.
We have now begun surveying different types of civic associations to compare
them to neighborhood councils. Unfortunately, these other groups are
informal and have no membership lists that we can employ for a saturation
sample. Consequently we were forced to rely on name generators to gather
Is there a way that we can correct for the inherent biases between these
sampling methods to make valid comparisons between the groups? Because the
neighborhood council data is used for other purposes, it is not possible to
employ name generators for those groups. Would it be justifiable to
"sample" from data collected through the saturation survey to employ only
six of the names identified for each inidividual (we collect up to six names
for the name generator). If so, are there better ways to "sample" than
others? We have data on the importance of links, so we could select the 6
most important alters nominated by each individual.
Alternatively, has anyone analyzed the biased created by differing sampling
methodologies? If all biases were known, we could at least account for them
in our comparisons.
Thank you in advance.
SOCNET is a service of INSNA, the professional association for social
network researchers (http://www.insna.org). To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.