Information Technology Reference
In-Depth Information
those co-branded on Amazon.com, and those
not available at the time of data collection. The
companies included in the Travel category were
identified as successful travel sites by the busi-
ness press (Ebenkamp, 2002; Forrester Research,
2002). The Web sites in the News and Internet
Services categories were selected intuitively in
an attempt to choose the most well-known Web
sites in these industries. The sample contains a
subsample of 19 companies that display one or
more privacy seals and, henceforth, are referred
to as seal companies .
preserve the original layout. Together, the privacy
policies of the 50 Web sites contained 108,570
words, with document lengths varying from
575 to 6,139 words. In a pilot study of the four
longest privacy policies in the sample, a coding
frame (see Appendix B) was developed, based
on an inductive coding strategy applied to these
four documents. Inductive coding refers to the
close and thorough study of documents in an
unrestricted manner to open up the inquiry and
to identify relevant categories that fit the data
(Strauss & Corbin, 1990). These inductively de-
rived codes were factual rather than referential in
nature. Factual codes condense the information
contained in a text by denoting precisely defined
facts, whereas referential codes represent only
themes in a text (Kelle & Laurie, 1995). In this
study, the factual codes referred to countable
features or were closed-ended questions referring
to data handling practices with the code values
yes , no , or no answer .
After a pilot coding of five other privacy poli-
cies and subsequent amendments to the coding
sheet, the coding frame consisted of 35 questions
pertaining to general characteristics (11 codes),
data collection (15 codes), marketing communica-
tions (one code), and data sharing (eight codes).
The coding was based on the at-least-some rule,
which considers practices true, even if they are
carried out only occasionally. All 50 privacy
statements were coded by the author in July 2003.
Of the 35 questions applied to the 50 privacy
policies, no answers were found to 22.86% of
these questions; thus, their code values were no
answer . This level was 19.55% for the seal compa-
nies and 24.88% for non-seal companies. Taking
into account only those 24 questions focusing on
specific data handling practices (see codes 12-35,
Appendix B), these levels were 33.33% for the
total sample, 28.51% for the seal companies, and
36.29% for the non-seal companies.
Clearly, the type of metric used has a bearing
on the types of analyses that may be conducted
(Bauer, 2000). The nominal scales used in the
methods of analysis
To analyze and compare what data handling
practices the sample companies engaged in or did
not engage in, a content analysis of the 50 privacy
policies was conducted. The goal of content analy-
ses was to examine message characteristics in an
objective manner by applying consistent criteria
rigorously (Neuendorf, 2002). The documents
analyzed represent how the writer “naturally
presents or projects itself to its normal audience,
rather than to a researcher” (Kabanoff, 1996, p.
6). Essentially, content analysis reduces data by
categorizing content into predefined categories,
thereby quantifying qualitative data. The results
obtained from content analyses are quantitative
indices of textual content and, thus, categorical
and descriptive in nature. They show not only
what is in a text but also what is not (Kolbe &
Burnett, 1991). Other analytic approaches, such
as discourse analysis or critical theory, concen-
trate on what is said and how it is said but do not
systematically assess what is not said (Lacity
& Janson, 1994). Content analysis was chosen
for this study, because its major strengths (i.e.,
the systematic analysis of textual data and the
unobtrusive examination of naturally occurring
data) prevent biases in the collection and analysis
of the data.
To facilitate the analysis, each privacy policy
was saved as a text file and as an image file to
Search WWH ::




Custom Search