Secondary Data Analysis: A Method of which the Time Has Come

This paper discusses the viability of secondary data analysis as a research method in library and information science. It presents an illustrative research application utilizing secondary data analysis in the study of U.S. school librarians and technology integration.

Qualitative and Quantitative Methods in Libraries (QQML) 3:619 –626, 2014
_________________
Received: 21.4.2013 / Accepted: 12.12.2013 ISSN 2241-1925
© ISAST
Secondary Data Analysis:
A Method of which the Time Has Come
Melissa P. Johnston, PhD
School of Library and Information Studies, University of Alabama, Tuscaloosa, AL,
USA
Abstract: Technological advances have led to vast amounts of data that has been
collected, compiled, and archived, and that is now easily accessible for research. As a
result, utilizing existing data for research is becoming more prevalent, and therefore
secondary data analysis. While secondary analysis is flexible and can be utilized in
several ways, it is also an empirical exercise and a systematic method with procedural
and evaluative steps, just as in collecting and evaluating primary data. This paper asserts
that secondary data analysis is a viable method to utilize in the process of inquiry when a
systematic procedure is followed and presents an illustrative research application
utilizing secondary data analysis in library and information science research.
Keywords: secondary data analysis, school librarians, technology integration
1. Introduction
In a time where vast amounts of data are being collected and archived by
researchers all over the world, the practicality of utilizing existing data for
research is becoming more prevalent (Andrews, Higgins, Andrews, Lalor, 2012;
Schutt, 2011; Smith, 2008; Smith et al., 2011). Secondary data analysis is
analysis of data that was collected by someone else for another primary purpose.
The utilization of this existing data provides a viable option for researchers who
may have limited time and resources. Secondary analysis is an empirical
exercise that applies the same basic research principles as studies utilizing
primary data and has steps to be followed just as any research method. This
paper asserts that secondary data analysis is a viable method to utilize in the
process of inquiry when a systematic process is followed. This paper contributes
to the discussion of secondary data analysis as a research method for library and
information science (LIS) and utilizes a study of U.S. school librarians to
describe and illustrate the process, benefits, and limitations in conducting an
investigation utilizing secondary data analysis method.

Melissa P. Johnston620
2. Defining Secondary Data Analysis
The concept of secondary data analysis first emerged with Glaser’s discussion
of re-analyzing data “which were originally collected for other purposes” (1963,
p. 11), yet there remains a dearth of literature that specifically addresses the
process and challenges of conducting secondary data analysis research
(Andrews et al., 2012; Smith, 2008). Hakim (1982) defines secondary analysis
as “any further analysis of an existing dataset which presents interpretations,
conclusions or knowledge additional to, or different from, those presented in the
first report on the inquiry as a whole and its main results” (p. 1). Most research
begins with an investigation to learn what is already known and what remains to
be learned about a topic through reviewing secondary sources and investigations
others have previously conducted in the specified area of interest. Secondary
data analysis takes this one step further, including a review of previously
collected data in the area of interest. While secondary data analysis is a flexible
approach and can be utilized in several ways, it is also an empirical exercise
with procedural and evaluative steps, just as there are in collecting and
evaluating primary data (Doolan & Froelicher, 2009). Secondary data analysis
remains an under-used research technique in many fields, including LIS. Given
the increasingly availability of previously collected data to researchers, it is
important to further define secondary data analysis as a systematic research
method. Yet, few frameworks are available to guide researchers as they conduct
secondary data analysis (Andrews et al., 2012; Smith et al., 2011).
3. Process of Secondary Analysis
In conducting research, the area of investigation and the research questions
determine the method that the researcher follows. The research method consists
of how the researcher collects, analyzes, and interprets the data in the study
(Creswell, 2009). Secondary analysis is a systematic method with procedural
and evaluative steps, yet there is a lack of literature to define a specific process,
therefore this paper proposes a process that begins with the development of the
research questions, then the identification of the dataset, and thorough
evaluation the dataset. This procedure is illustrated by a LIS research study in
which the researcher investigated school librarians as leaders in technology
integration.
3.1 Develop the Research Questions
The key to secondary data analysis is to apply theoretical knowledge and
conceptual skills to utilize existing data to address the research questions.
Hence, the first step in the process is to develop the research questions. The
purpose of this study was to investigate the enablers and barriers that school
librarians experience enacting a leadership role in technology integration. The
research questions that guided this work are: What enablers or supporting
factors do accomplished school librarians perceive as enablers in enacting the
role of leader in technology integration? What barriers or constraining factors do
accomplished school librarians perceive to enacting the role of leader in

Qualitative and Quantitative Methods in Libraries (QQML) 3:619 –626, 2014 621
technology integration? What is the association between accomplished school
librarians involved at a high level in technology integration leadership and the
identified enablers/barriers in comparison to the other participants?
3.2 Identifying the Dataset
Most research begins with an investigation to learn what is already known and
what remains to be learned about a topic (Creswell, 2009); including related and
supporting literature, but one should also consider previously collected data on
the topic (Dale, Arbor, & Procter, 1988; Doolan & Froelicher, 2009). Data may
already exist that can be utilized in addressing the research questions.
In the case of this research an in-depth literature review of the areas of interest
was conducted examining the previous and current work of experts in the field
of school librarianship and technology. Through the literature review other
researchers on this topic were identified, as were agencies and research centers
that have conducted related studies. Recent research and findings from the top
ranked school library preparation programs were identified and reviewed, as
were dissertations in the areas of technology, leadership, and school librarians.
Finally, local informal networks can also provide valuable information in
determining what research is currently being conducted (Magee, Lee, Giuliano,
& Munro, 2006). This is especially relevant in the field of school librarianship,
which is a very connected community. The researcher had the benefit of an
informal network, in that she was a part of the team that worked on background
research and survey construction for a then current study by the Partnerships for
Advancing Library Media (PALM) Center at Florida State University (FSU).
Original survey research rarely uses all of the data collected and this unused
data can provide answers or different perspectives to other questions or issues
(Heaton, 2008, Johnston, 2012; Smith, 2008), yet the key to using existing
survey data effectively to find meaningful answers is a good fit between the
research question and the dataset (Doolan & Froelicher, 2009; Kiecolt &
Nathan, 1985; Magee et al, 2006). In this study, the research questions fit well
with that of the original study since both studies focused on school librarians
and technology leadership. The researcher’s relationship with the primary
investigators, made her aware that data collected from questions addressing
enablers and barriers to leadership in technology integration had not been
analyzed or reported. Finding that this data would adequately address her
research questions and that the primary method of data collection was
appropriately suited to her research, the decision was made to utilize existing
survey data to find the answers to different research questions than were asked
in the original research.
3.3 Evaluating the Dataset
Once a dataset that appears viable in addressing initial requirements discussed
above is located, the next step in the process is evaluation of the dataset to
ensure the appropriateness for the research topic (Dale et al., 1988; Kiecolt &
Nathan, 1985; Smith, 2008; Stewart & Kamins, 1993). The advantage is that the

Melissa P. Johnston622
data already exist in some form and can be evaluated for appropriateness and
quality in advance of actual use (Stewart & Kamins, 1993). Stewart and Kamins
(1993) propose a reflective approach to evaluate the data in a “stepwise fashion”
(p. 18). The following evaluative steps should be followed in order to determine
the appropriate match of a dataset to a research investigation and ensure
congruency, quality of the primary study and the resulting dataset: (a) what was
the purpose of this study; (b) who was responsible for collecting the
information; (c) what information was actually collected; (d) when was the
information collected; (e) how was the information obtained; and (f) how
consistent is the information obtained from one source with information
available from other sources (Stewart & Kamins, 1993). The researcher was
given access to and utilized all documentation on the collection of the data,
information found in publication, and consulted the investigators from the
primary study in order to complete this evaluation.
3.3.1 What was the purpose of this study? It is important to determine the
purpose of the original project that produced the data because this can influence
many factors such as the targeted population, the sample selected, the wording
of questions on the survey, and the general context of the study (Doolan &
Froelicher, 2009; Magee et al., 2006). It is also important to know about the
agency or individual(s) that collected the information and the similarities or
differences in research goals between those researchers and the researcher
contemplating secondary analysis (Boslaugh, 2007; Stewart & Kamins, 1993).
The researcher had the benefit of a professional relationship with the researchers
who conducted the original study and through maintaining contact was privy to
inside information about the data collection process. The original study sought
to characterize the technology integration activities of school librarian leaders in
order to answer the research question: “What is the leadership role of the school
librarian in technology integration?” The overall research goal to improve the
education of future school librarians coincides with the secondary researcher’s
goal. However, in this case secondary data analysis allows the researcher to
analyze the original dataset to answer a different question; in this case, what is
enabling or deterring these practices.
3.3.2 Who was responsible for collecting the information? In addressing the
question of who was responsible for collecting the information the secondary
researcher again has the benefit of a relationship with the primary research team.
However, she conducted an investigation into the primary investigators’
backgrounds and previous research projects finding that that the primary
investigators are well-respected academic researchers, and have a reputation for
excellence in research integrity.
3.3.3 What information was actually collected? It is vital for the secondary
researcher to have access to adequate documentation from the primary research,
including protocols and procedures followed in the collection of the data (Clarke
& Cossette, 2000; Dale et al., 1988; Smith, 2008; Smith et al., 2011; Stewart &
Kamins, 1993). The primary research team kept detailed documentation that

Qualitative and Quantitative Methods in Libraries (QQML) 3:619 –626, 2014 623
provided evidence of careful and consistent data collection. The documentation,
the survey instrument, and published findings were consulted, finding that the
survey consists of three sections and collected the following data: 30
demographic questions; 60 Likert scale statements on technology integration
practices; and three open-ended questions that asked respondents to discuss
barriers, enablers, and other factors that influenced their leadership practices
(Everhart, Mardis, & Johnston, 2012).
3.3.4 When was the information collected? In any research the time when the
data is collected must be considered (Boslaugh, 2007; Stewart & Kamins,
1993). Survey data may be several years old before it is released and available
for use by others. In areas related to technology, as in this research, the time
frame of data collection is paramount (Boslaugh, 2007). The researcher looked
for data related to school librarians as technology leaders that were no more than
three years old. The FSU researchers conducted the primary study and collected
data during the spring, summer, and fall of 2009, making this data at that time,
the most current dealing with this topic.
3.3.5 What methodology was employed in obtaining the data? “The quality
of secondary data cannot be evaluated without knowledge of the methodology
employed when collecting the data” (Stewart & Kamins, 1993, p. 25). The
secondary researcher must be knowledgeable of the primary method as well. In
evaluating the existing data, issues with the survey method including the survey
instrument had to be considered. The primary research team developed the
School Librarian Technology Leadership Survey (PALM, 2009), because no
instrumentation existed in this area. One disadvantage of utilizing secondary
data is that secondary researchers often have to settle for the original
measurement tool and therefore have to evaluate and make a judgment call on
the instrumentation (Clarke & Cossette, 2000). Even though the researcher was
involved in constructing the survey instrument in the primary research, the
literature review and documentation of this process were consulted to address
validity and reliability issues (Magee et al., 2006). Finally, in evaluating how
the original data was collected, the researcher examined how issues such as
sampling, response rates, missing responses, and bias were handled in the
original research (Kiecolt & Nathan, 1985).
3.3.6 Management of the primary data. It is mandatory for the secondary
analyst researcher to obtain all documentation of the processes and protocols
followed by the primary researchers, including the questionnaire, all coding
materials, and any publications that are related to the data (Boslaugh, 2007;
Clarke & Cossette, 2000; Stewart & Kamins, 1993). Finally, it is paramount that
the secondary researcher has access to the raw dataset in order to perform new
analyses and to consider and account for all of the aforementioned possible
concerns (Boslaugh, 2007; Stewart & Kamins, 1993). The researcher was
granted permission to access to the raw dataset and all supporting
documentation.

Melissa P. Johnston624
3.3.7 How consistent is the information obtained from one source with
information available from other sources? It is beneficial to have multiple
sources to bolster confidence in findings, whether it is that two or more sources
arrive at the same conclusion for comparison or that they do not, providing an
option for contrast. In the case of this research, there was no other similar data
located on this very specialized topic, so this analysis was impossible.
4. Discussion: Strengths and Limitations of Secondary Analysis for
LIS Research
The major advantages associated with secondary analysis are the cost-
effectiveness and convenience it provides (Dale et al., 1988; Glaser, 1962;
Smith, 2008). Since someone else has already collected the data, the researcher
does not have to devote financial resources to the collection of data. When good
secondary data is available, researchers can gain access to and utilize high
quality larger datasets, such as those collected by funded studies or agencies that
involve larger samples and contain substantial breadth. The larger samples are
more representative of the target population and allow for greater validity and
more generalizable findings (Smith, 2008; Smith et al., 2011). Access to this
type of data presents opportunities for all researchers, even the novice or
unfunded researcher, therefore equalizing opportunities and building capacity
for empirical research (Hakim, 1982) in LIS research.
The use of existing data sets can accelerate the pace of research because some of
the most time consuming steps of a typical research project, such as
measurement development and data collection are eliminated (Doolan &
Froelicher, 2009). In LIS research areas, such as information and technology
that are constantly changing, utilizing existing data allows projects to be
completed and findings to be produced much faster, and therefore the
development and contribution of new knowledge occurs in a timely manner
before they are considered dated by the field. Additionally, in the area of
information policy, utilizing existing data can allow the researcher to answer
important time-sensitive policy related questions quicker (Magee et al., 2006).
Secondary data analysis provides many opportunities for furthering LIS research
through replication, re-analysis and re-interpretation of existing research. It
provides researchers with opportunities to engage in work to test new ideas,
theories, frameworks, and models of research design.
Yet there are unique methodological considerations when utilizing existing data
to investigate new research questions and generate new knowledge. The most
recognized limitation to the secondary data analysis method approach is
“inherent in its nature” in that the data were collected for some other purpose
(Boslaugh, 2007, p. 4). Since the data were not collected to answer the
researcher’s specific research questions issues can arise. The specific
information that the researcher would like to have may not have been collected;
or data may not have been collected in the geographic region of interest, in the
years the researcher would have chosen, or on the specific population that is the

Qualitative and Quantitative Methods in Libraries (QQML) 3:619 –626, 2014 625
focus of interest (Boslaugh, 2007; Doolan & Froelicher, 2009).
In this particular project the researcher avoided some common pitfalls often
associated with secondary analysis by participating in the primary research
design plan and then ensuring a match between her research questions and the
existing data through the previously described process. Yet, a significant
limitation of this research was that the school identifiers collected in the primary
study were not available to the researcher due to confidentiality reasons. The
school identifiers connect to the participants, therefore school identifiers were
removed from the dataset, in order to ensure all participants remain anonymous
in accordance with the original consent agreement. Therefore, subjects cannot
be contacted for follow-up questions and additional data cannot be collected.
While this lack of opportunity for follow-up or the collection of additional data
from the participants has proven to be a limitation in furthering this research, it
is important that secondary data analysis abide by the consent conditions of the
original study (Heaton, 2008).
A second major disadvantage of using secondary data is that the secondary
researcher did not participate in the data collection process and does not know
exactly how it was conducted. Therefore, the secondary researcher does not
know how well it was done and if the data are affected by problems such as low
response rate or respondent misunderstanding of specific survey questions.
Hence the researcher has to find this information through other means such as
documentation of the data collection procedures, technical reports, and
publications (Boslaugh, 2007; Dale et al., 1988; Kiecolt & Nathan, 1985). In
this research the researcher was at a disadvantage because she did not
participate in the execution of the data collection process. In order to address
these issues the researcher utilized documentation from the original study,
information from published findings, and consultations with the original
primary researchers and statistician. Ensuring a match between the research
question and the existing data and following a process, as proposed, for careful
reflective examination and critical evaluation of the data, can avoid most
limitations of secondary data analysis.
5. Conclusion
Secondary data analysis offers methodological benefits and can contribute to
LIS research through generating new knowledge (Heaton, 2008, Johnston, 2012;
Smith, 2008). The overall goal of this method is the same as that of others, to
contribute to scientific knowledge through offering an alternate perspective; it
only differs in its reliance on existing data. LIS researchers should take
advantage of the high quality data that are available and consider the potential
value in gaining knowledge and giving insight into a broad range of LIS issues
through utilizing secondary data analysis method.
Yet, successful secondary analysis of data requires a systematic process that
acknowledges challenges of utilizing existing data and addresses the distinct
characteristics of secondary analysis. The process proposed from this

Melissa P. Johnston626
application in LIS research provides a systematic process that includes steps to
undertake to avoid possible limitations. In a time where the large amounts of
data being collected, compiled, and archived by researchers all over the world
are now more easily accessible, the time has definitely come for secondary data
analysis as a viable method for LIS research.
References
Andrews, L., Higgins, A., Andrews, M. W., & Lalor, J. G. (2012). Classic grounded
theory to analyse secondary data: Reality and reflections. The Grounded Theory
Review, 11(1), 12-26.
Boslaugh, S. (2007). Secondary analysis for public health: A practical guide. New
York, NY: Cambridge. doi: 10.1017/CBO9780511618802
Clarke, S. P., & Cossette, S. (2000). Secondary analysis: Theoretical, methodological
and practical considerations. Canadian Journal of Nursing Research, 32(3), 109-129.
Creswell, J. W. (2009). Research design: Qualitative, quantitative, and mixed
methods approaches (3rd ed.). Thousand Oaks, CA: Sage.
Dale, A., Arbor, S., & Procter, M. (1988). Doing secondary analysis. London, UK:
Unwin Hyman.
Doolan, D. M., & Froelicher, E. S. (2009). Using an existing data set to answer new
research questions: A methodological review. Research and Theory for Nursing
Practice: An International Journal, 23(3), 203-215. doi:10.1891/1541-6577.23.3.203
Everhart, N., Mardis, M., & Johnston, M. P. (2012). National Board Certified school
librarians’ leadership in technology integration: Results of a national survey. School
Library Media Research, 14.
Glaser, B. G. (1963). Retreading research materials: The use of secondary analysis by
the independent researcher. The American Behavioural Scientist, 6(10), 11-14.
Hakim, C. (1982). Secondary analysis in social research: A guide to data sources
and method examples. London, UK: George Allen & Uwin.
Heaton, J. (2008). Secondary analysis of qualitative data: An overview. Historical
Social Research, 33(3), 33-45.
Johnston, M. P. (2012). School librarians as technology integration leaders: Enablers
and barriers to leadership enactment. School Library Research, 15(1).
Kiecolt, K. J., & Nathan, L. E. (1985). Secondary analysis of survey data. Sage
University Paper Series on Quantitative Applications in the Social Sciences, 53.
Magee, T., Lee, S. M., Giuliano, K. K., & Munro, B. (2006). Generating new
knowledge from existing data: The use of large data sets for nursing research. Nursing
Research, 55(2), S50-S56.
Partnerships Advancing Library Media (PALM) Center, Florida State University
(2009). School library media specialist and technology integration survey. Unpublished
Smith, E. (2008). Using secondary data in educational and social research. New
York, NY: McGraw-Hill Education.
Smith, A. K., Ayanian, J. Z., Covinsky, K. E., Landon, B. E., McCarthy, E. P., Wee,
C. C., & Steinman, M. A. (2011). Conducting high-value secondary dataset analysis: An
introductory guide and resources. Journal of General Internal Medicine, 28(8), 920-
929. doi:10.1007/s11606-010-1621-5
Stewart, D. W., & Kamins, M. A. (1993). Secondary research: Information sources
and methods. Newbury Park, CA: Sage.