Using Secondary Data in
Doctoral Research
By Jim Goes and
Marilyn Simon
When considering research topics and research designs, most
doctoral students decide to collect primary data. Primary data are data that are proactively
gathered for a specific research purpose, and both the choice of
instrumentation for data collection and use of the resulting data are focused
on addressing the specific purpose and the alignment among the research
question, planned analyses, and types of data collection proposed (Simon &
Goes, 2013) The most attractive feature of primary data is that they are
customized to the particular study being completed, and available for use and analysis once gathered.
There are, however, challenges involved with gathering
primary data.. Not only must a population
and sample be identified clearly, but as researchers we must convince our
target population to respond. People are
constantly surveyed, interviewed, asked for feedback, etc., and often resist
the time and effort necessary to participate in market or academic
research.
Regardless of whether interviews, surveys, focus groups, or
some other approach is being used, one of the most difficult problems facing
researchers is gaining an acceptable level of response to enable the planned
analyses. This may be less difficult
with some qualitative designs, like phenomenology, because the necessary sample
size is usually smaller. In many cases,
however, students end up with fewer responses than needed, and must adapt their
analyses and statistical approach to account for a less than ideal sample size.
There is
another approach, and it can be quite effective even though among doctoral
students it is rarely used. Few students
take advantage of secondary sources of
data, even when they are readily available.
Secondary data consists of data that were collected for a different
purpose, but can be repurposed for use in a different study. These data may be publicly available for students
to use, such as census data, statistical agencies, federal agencies, academic
publications, and. trade associations. There are also data considered internal
to an organization that might be obtained with permission, such as standardized
tests in an educational institution, or annual reports from businesses, which
may require a royalty or payment to use the data. Data from national, state, and local
governments is often available free of charge, although it may take some work
to access. In some cases, like that of
census data, large portions of the data may even be downloaded off the
Internet.
For example, I (Jim) completed my dissertation years ago
based on secondary data obtained from the State of California. The data consisted of profile, operational,
and business data on California hospitals.
Hospitals were required to gather these data and submit it to the state
on an annual basis as part of the state regulatory framework. Even better, the data were required to be
audited, providing strong assurance of quality.
After some digging, I found that the data were available from the state
just for the asking. What a deal! Even though it took some work, a lot of phone
calls, and even a trip to Sacramento (state capitol) to obtain the data, it was
well worth it.
The core challenge in using secondary data usually revolves
around finding ways to make the data fit with the purpose of the new study and
the research questions being asked. In
my case, I was studying processes of strategic and structural change in
hospitals over a period of a dozen years.
The unit of analysis was the organization (hospital), and very few
researchers in my field had access to any sort of longitudinal,
organization-level data like this. The
real work in using this data was in figuring out how to adapt and apply it to
fit my research goals, and how to craft variables and measures that were
credible indicators of the kinds of phenomena I was interested in
studying. In some cases, I had to
transform the data, create constructs and measures by combining multiple indicators,
and overall be very inventive coming up with measures that served as proxies
for the actual variables I wanted to study.
In the end, it was all worth it. I was not required to gather any primary
data, which meant no turn downs by participants, no missing data, and no
pleading e-mails to members of the study population asking them to help me.
However I did spend far more time on preparing, cleaning, and recoding the
secondary data to fit my purposes than I would have with a primary study. The result, however, was a database and a study
that was unusual in the field and had scope and breadth far beyond most primary
studies, enabling me to answer questions that had not yet been investigated
fully because of the absence of primary data on the topic.
So as you
are considering the methodology and design for your dissertation or other
research study, consider secondary data.
Rather than gravitating immediately to a strategy that involves primary
data collection, think about how you can gain access to secondary data, already
collected and available, that can address your research questions. You may even find that a mixture of primary
and secondary data will make for a stronger study and faster time to completion
of your study or dissertation.
References
Cowton,
C. J. (1998). The use of secondary data in business ethics research. Journal of Business Ethics, 17(4), 423-434.
Huston,
P., & Naylor, C. D. (1996). Health services research: reporting on studies
using secondary data sources. CMAJ:
Canadian Medical Association Journal, 155(12),
1697.
Nicoll,
L. H., & Beyea, S. C. (1999). Using secondary data analysis for nursing
research. AORN journal, 69 (2),
428-428.
Simon,
M. & Goes, J (2013).
Szabo,
V., & Strang, V. R. (1997). Secondary analysis of qualitative data. Advances in Nursing Science, 20(2), 66-74.
Smith, E.
(2008). Using secondary data in educational and social research. Open
University Press.
sumber : http://dissertationrecipes.com/wp-content/uploads/2011/04/Using-Secondary-Data-in-Doctoral-Research.pdf
Tidak ada komentar:
Posting Komentar