[Air-L] Methodology Question Help

Nikki Usher nusher at usc.edu
Tue May 24 15:30:43 PDT 2011


Hiya:

I don't think this got picked up?

---------- Forwarded message ----------
From: Nikki Usher <nikki.usher at gmail.com>
Date: Mon, May 23, 2011 at 12:03 PM
Subject: Methodology Question Help
To: air-l at listserv.aoir.org


Hi:

Normally I recruit subjects for interviews in person, but this time I'm
using email and I'm having a little trouble figuring out what the best
practices are for methodology. Let me explain about how we got our sample
without giving TOO many details since one of you might be a kind reviewer.
As you may have gathered by now, I write about new media and journalism.

1) An organization provided us a list of news articles with just their
headlines. 84 articles in total. So we had to find the bylines, emails and
contact information. journalists are typically regarded as public figures by
the IRB on account of their publishing activities and public persona. The
people we were able to find and contact all had Web sites promoting their
work with email addresses, etc.

2) We recruited participants via email using IRB approved language (obvs) -
we were able to get approx. 34 email addresses from this initial data set -
and so far have 15 respondents.  We have promised to keep their responses
anonymous, though the "star" of the effort is now a known quantity and would
be identified as such by her responses, so she's OK with us using her name.

3) We are almost close to saturation with the thematic interview responses
but we'd like more interviews to be sure. We also think a potential reviewer
would think "15, really, that's super small." We are conducting some follow
up emails with this first batch of email addresses. Our concern here is
self-selection of respondents. However, isn't self-selection  always the
concern with interviews (Weiss - talking to strangers, Silverman, etc.
etc.)?  I just don't know what we should do about response rate because I
don't normally recruit over email to a widely dispersed group of people who
don't have any connection to each other. In a real life organization, you
might say, talk to 15 of the reporters who all have something in common that
you've been able to observe, or select to interview people based on
hierarchy.

4) We have *another *data set where if we can't get saturation we can look
for participants but it is much larger (about 150) - from a second wave of
the organization's story output. Would we need to send out emails to all of
these journalists to remain consistent, or could we just approach the first,
say, dozen, going by date of story publication until we got saturation? Or
would we have to start over and email EVERYONE? We are pretty sure we will
reach saturation of themes quite soon.

Thanks!!

-- 
Nikki Usher, PhD
USC Annenberg School for Communication & Journalism
mobile: 213-220-7824
www.nikkiusher.com



More information about the Air-L mailing list