Week 6 Potential Theoretical Data Analysis and Verification Plan Homework Based on the data collection plan developed in week 6, develop the data analysis and verification plan. Write 5 pages paper that includes the following:Research questions (Sources of data: Refer to Onwuegbuzie and Leech (2006) in week-4 readings, and describe which of the data analysis steps outlined on page 491 and 492 (Figure 3) will be followed. Refer to Adams-Budde at al. (2014) in week 7 readings to describe how each source of data will be used in the analysis. Use other readings as models to discuss data integration procedures.Refer to articles provided in weeks 7 and 8 to discuss specific verification strategies for proposed data analysis: how will the issues of validity (quantitative data) and trustworthiness (qualitative data) be addressed? Make sure to cite articles from the readings to support your procedures.Discuss limitations of the proposed data analysis plan and reflect on the potential theoretical and practical significance of your work. International Journal of Education
ISSN 1948-5476
2011, Vol. 3, No. 1: E13
Data Analysis in Mixed Research: A Primer
Anthony J. Onwuegbuzie (Corresponding author) & Julie P. Combs
Department of Educational Leadership and Counseling
Sam Houston State University, USA
Box 2119, Sam Houston State University, Huntsville, Texas 77341-2119
E-mail: tonyonwuegbuzie@aol.com
Received: March 23, 2011
Accepted: April 20, 2011
doi:10.5296/ije.v3i1.618
Abstract
The purpose of this methodological article is to provide a primer for conducting a mixed
analysis—the term used for analyzing data in mixed research. Broadly speaking, a mixed
analysis involves using quantitative and quantitative data analysis techniques within the same
study. In particular, a heuristic example using real data from a published study entitled
“Perceptions of Barriers to Reading Empirical Literature: A Mixed Analysis” (Benge,
Onwuegbuzie, Burgess, & Mallette, 2010) is used with the aid of screenshots to illustrate
how a researcher can conduct a quantitative dominant mixed analysis, wherein the
quantitative analysis component is given higher priority and qualitative data and analysis is
incorporated to increase understanding of the underlying phenomenon.
Keywords: Mixed research, Mixed methods research, Quantitative research, Qualitative
research, Mixed analysis, Analysis screenshots
1
www.macrothink.org/ije
International Journal of Education
ISSN 1948-5476
2011, Vol. 3, No. 1: E13
1. Mixed Research Defined
Mixed Research, or what is referred to as mixed methods research, involves “mix[ing] or
combin[ing] quantitative and qualitative research techniques, methods, approaches, concepts
or language into a single study” (Johnson & Onwuegbuzie, 2004, p. 17). As noted by Collins,
Onwuegbuzie, and Sutton (2006), mixed research studies contain 13 steps—each of which
occur at one of the following three phases of the mixed research process: research
conceptualization (i.e., determining the mixed goal of the study, formulating the mixed
research objective[s], determining the rationale of the study and rationale[s] for mixing
quantitative and qualitative approaches, determining purpose of the study and the purpose[s]
for mixing quantitative and qualitative approaches, determining the mixed research
question[s]), research planning (i.e., selecting the mixed sampling design, selecting the mixed
research design), and research implementation (i.e., collecting quantitative and qualitative
data, analyzing the quantitative and qualitative data, legitimating the data sets and mixed
research findings, interpreting the mixed research findings, writing the mixed research report,
reformulating the mixed research question[s]). Of these 13 steps, analyzing data in a mixed
research study potentially is the most complex step because the researcher(s) involved has to
be adept at analyzing both the quantitative and qualitative data that have been collected, as
well as integrating the results that stem from both the quantitative and qualitative analysis “in
a coherent and meaningful way that yields strong meta-inferences (i.e., inferences from
qualitative and quantitative findings being integrated into either a coherent whole or two
distinct sets of coherent wholes; Tashakkori & Teddlie, 1998)” (Onwuegbuzie & Combs,
2010, p. 398). As such, guidelines and exemplars are needed for conducting mixed analyses.
Thus, the purpose of this article is to describe and to illustrate data in mixed research.
2. Mixed Analysis Defined
Mixed analysis is the term used for analyzing data in mixed research. Onwuegbuzie and
Combs (2010) recently provided an inclusive definition of mixed analysis that incorporates
the definition and typologies that have been presented in major methodological works. These
works included articles, book chapters, books, and paper presentations across numerous fields
and disciplines such as the social and behavioral sciences (including psychology and
education), nursing and allied health, business, and linguistics that spanned 21 years. Based
on their interpretations of the extant literature, Onwuegbuzie and Combs (2010) identified 13
criteria that represent decisions that mixed researchers make before, during, and/or after the
conduct of their mixed analyses:
1. rationale/purpose for conducting the mixed analysis
2. philosophy underpinning the mixed analysis
3. number of data types that will be analyzed
4. number of data analysis types that will be used
5. time sequence of the mixed analysis
6. level of interaction between quantitative and qualitative analyses
7. priority of analytical components
8. number of analytical phases
2
www.macrothink.org/ije
International Journal of Education
ISSN 1948-5476
2011, Vol. 3, No. 1: E13
9. link to other design components
10. phase of the research process when all analysis decisions are made
11. type of generalization
12. analysis orientation
13. cross-over nature of analysis
Using these 13 criteria, Onwuegbuzie and Combs (2010) derived the following inclusive and
comprehensive definition of mixed analysis:
Mixed analysis involves the use of both quantitative and qualitative analytical
techniques within the same framework, which is guided either a priori, a posteriori, or
iteratively (representing analytical decisions that occur both prior to the study and
during the study). It might be based on one of the existing mixed methods research
paradigms (e.g., pragmatism, transformative-emancipatory) such that it meets one of
more of the following rationales/purposes: triangulation, complementarity,
development, initiation, and expansion. Mixed analyses involve the analysis of one or
both data types (i.e., quantitative data or qualitative data; or quantitative data and
qualitative data), which occur either concurrently (i.e., in no chronological order), or
sequentially in two phases (in which the qualitative analysis phase precedes the
quantitative analysis phase or vice versa, and findings from the initial analysis phase
inform the subsequent phase) or more than two phases (i.e., iteratively). The analysis
strands might not interact until the data interpretation stage yielding a basic parallel
mixed analysis, although more complex forms of parallel mixed analysis can be used,
in which interaction takes place in a limited way before the data interpretation phase.
The mixed analysis can be designed based, wherein it is directly linked to the mixed
methods design (e.g., sequential mixed analysis techniques used for sequential mixed
methods designs). Alternatively, the mixed analysis can be phase based, in which the
mixed analysis takes place in one or more phases (e.g., data transformation). In mixed
analyses, either the qualitative or quantitative analysis strands might be given priority
or approximately equal priority as a result of a priori decisions (i.e., determined at the
research conceptualization phase) or decisions that emerge during the course of the
study (i.e., a posteriori or iterative decisions). The mixed analysis could represent
case-oriented, variable-oriented, and process/experience oriented analyses. The mixed
analysis is guided by an attempt to analyze data in a way that yields at least one of
five types of generalizations (i.e., external statistical generalizations, internal
statistical generalizations, analytical generalizations, case-to-case transfer, naturalistic
generalization). At its most integrated form, the mixed analysis might involve some
form of cross-over analysis, wherein one or more analysis types associated with one
tradition (e.g., qualitative analysis) are used to analyze data associated with a different
tradition (e.g., quantitative data). (pp. 425-426)
Of these 13 decision criteria, the following five criteria appear to be most common: (a)
rationale/purpose for conducting the mixed analysis, (b) number of data types that will be
analyzed, (c) time sequence of the mixed analysis, (d) priority of analytical components, and
(e) number of analytical phases. Each of these criteria is described in the subsequent sections.
3
www.macrothink.org/ije
International Journal of Education
ISSN 1948-5476
2011, Vol. 3, No. 1: E13
Rationale/purpose for conducting the mixed analysis
Greene, Caracelli, and Graham (1989) identified five purposes for mixing quantitative and
qualitative data: triangulation (i.e., quantitative findings are compared to the qualitative
results); complementarity (i.e., results from one analysis type [e.g., qualitative] are interpreted
to enhance, expand, illustrate, or clarify findings derived from the other strand [quantitative]);
development (i.e., data are collected sequentially and the findings from one analysis type are
used to inform data collected and analyzed using the other analysis type); initiation (i.e.,
contradictions or paradoxes that might reframe the research question are identified), and
expansion (i.e., quantitative and qualitative analyses are used to expand the study’s scope and
focus).
Number of data types that will be analyzed
Traditionally, as noted by Creswell and Plano Clark (2007), “Data analysis in mixed methods
research consists of analyzing the quantitative data using quantitative methods and the
qualitative data using qualitative methods” (p. 128). However, mixed analyses also can
involve the sequential analysis of one data type—which are referred to as sequential mixed
analyses (Tashakkori & Teddlie, 1998), wherein data that are generated from the initial
analysis then are converted into the other data type. For example, a researcher could conduct
a qualitative analysis of qualitative data followed by a quantitative analysis of the qualitative
codes that emerge from the qualitative analysis and that are transformed to quantitative data
(e.g., exploratory factor analysis of themes that emerge from a constant comparison analysis
of qualitative data; cf. Onwuegbuzie, 2003). Such conversion of qualitative data into
numerical codes that can be analyzed quantitatively (i.e., statistically) is known as
quantitizing (Miles & Huberman, 1994; Tashakkori & Teddlie, 1998). Alternatively, a
researcher could conduct a quantitative analysis of quantitative data followed by a qualitative
analysis of the quantitative data that emerge from the quantitative analysis and that are
transformed to qualitative data (e.g., narrative profile formation of a set of test scores or
subscale scores representing the affective domain). Such conversion of quantitative data into
narrative data that can be analyzed qualitatively is known as qualitizing (Tashakkori &
Teddlie, 1998).
Time sequence of the mixed analysis
Time sequence refers to whether the quantitative and qualitative analysis components occur
in a chronological order (Creswell & Plano Clark, 2007). Specifically, the qualitative and
quantitative analyses can be conducted in chronological order, or sequentially (i.e., sequential
mixed analysis), or they can be conducted in no chronological order, or concurrently (i.e.,
concurrent mixed analysis). When sequential mixed analyses are conducted, either (a) the
quantitative analysis component is conducted first, which then drives or informs the
subsequent qualitative analysis component (i.e., sequential quantitative-qualitative analysis;
Onwuegbuzie & Teddlie, 2003); (b) the qualitative analysis component is conducted first,
which then informs the subsequent quantitative analysis component (i.e., sequential
qualitative-quantitative analysis; Onwuegbuzie & Teddlie, 2003); or (c) the quantitative and
qualitative analyses are conducted sequentially in more than two phases (i.e., iterative
sequential mixed analysis; Teddlie & Tashakkori, 2009).
4
www.macrothink.org/ije
International Journal of Education
ISSN 1948-5476
2011, Vol. 3, No. 1: E13
Priority of analytical components
Another important aspect of mixed analyses is the priority or emphasis given to the
quantitative analysis component(s) and the qualitative analysis component(s). Either the
qualitative and quantitative analysis components can be given approximately equal priority
(i.e., equal status) or one analysis component can be given significantly higher priority than
the other analysis component (i.e., dominant status). If the quantitative analysis component is
given significantly higher priority, then the analysis essentially is a quantitative-dominant
mixed analysis, wherein the analyst adopts a postpositivist stance, while believing
simultaneously that the inclusion of qualitative data and analysis is likely to increase
understanding of the underlying phenomenon (cf. Johnson, Onwuegbuzie, & Turner, 2007).
In contrast, if the qualitative analysis component is given significantly higher priority, then
the analysis essentially is a qualitative-dominant mixed analysis, whereby the analyst
assumes a constructivist-poststructuralist-critical stance with respect to the mixed analysis
process, while believing simultaneously that the inclusion of quantitative data and analysis is
likely to provide richer data and interpretations (cf. Johnson et al., 2007).
Number of analytical phases
Mixed analyses involve several phases. For example, Greene (2007, p. 155) identified the
following four phases of analysis: (a) data transformation, (b) data correlation and
comparison, (c) analysis for inquiry conclusions and inferences, and (d) using aspects of the
analytic framework of one methodological tradition within the analysis of data from another
tradition. Onwuegbuzie and Teddlie (2003) conceptualized a seven-step process for mixed
analyses: (a) data reduction (i.e., reducing the dimensionality of the quantitative data and
qualitative data), (b) data display (i.e., describing visually the quantitative data and qualitative
data), (c) data transformation (i.e., quantitizing and/or qualitizing data), (d) data correlation
(i.e., correlating quantitative data with quantitized data or correlating quantitative data with
qualitized data), (e) data consolidation (i.e., combining both quantitative and qualitative data
to create new or consolidated variables or data sets), (f) data comparison (i.e., comparing data
from the quantitative and qualitative data sources), and (g) data integration (i.e., integrating
both qualitative and quantitative data into a coherent whole).
Heuristic Example
The following mixed research study (Benge, Onwuegbuzie, Burgess, & Mallette, 2010)
provides an example of how one can conduct a mixed analysis. This study is relevant to any
field because it involves the study of reading ability within the context of doctoral-level
research methods courses.
Purpose of the Study
The purpose of Benge et al.’s (2010) study was fourfold: (a) to examine levels of reading
ability—as measured by reading comprehension and reading vocabulary—among doctoral
students; (b) to identify doctoral students’ perceptions of barriers that prevented them from
reading empirical articles; (c) to examine the relationship between these perceived barriers
and levels of reading vocabulary and reading comprehension; and (d) to determine which
5
www.macrothink.org/ije
International Journal of Education
ISSN 1948-5476
2011, Vol. 3, No. 1: E13
perceived barriers predict the perceived difficulty that doctoral students experience in reading
empirical research articles.
Participants were 205 doctoral students enrolled in one of the doctoral-level research design
courses at a large research university in the United States. Because all participants
contributed to both the qualitative and quantitative phases of the study, and the qualitative
and quantitative data were collected concurrently, the mixed sampling design used was a
Concurrent Design using Identical Samples (Onwuegbuzie & Collins, 2007). Although in the
study the quantitative and qualitative approaches were given approximately equal weight, the
researchers placed a greater emphasis on the quantitative analysis phase, yielding a
quantitative- dominant mixed analysis. The rationale/purpose for mixing quantitative and
qualitative analysis was complementarity and expansion (Greene et al., 1989).
All participants were administered the Nelson-Denny Reading Test (NDRT; Brown, Fishco,
& Hanna, 1993) and the Reading Interest Survey (RIS). The NDRT was used to measure
levels of reading vocabulary (80 items; KR-20 = .85) and reading comprehension (38 items;
KR-20 = .69). The RIS contains 62 items that are either open-ended (e.g., “What barriers
prevent you from reading more empirical research articles?”) or closed-ended (e.g., “Please
indicate your perceptions about the levels of ease/difficulty you experience in reading
empirical research articles. Please check the option that best applies: 1 = EASY; 2 =
SOMEWHAT EASY; 3 = NEITHER EASY NOR DIFFICULT; 4 = SOMEWHAT
DIFFICULT; 5 = DIFFICULT”). Figure 1 displays part of these data.
Quantitative Dominant Mixed Analysis: Stage-by-Stage
A sequential mixed analysis (SMA; Onwuegbuzie & Teddlie, 2003; Tashakkori & Teddlie,
1998) was conducted to analyze doctoral students’ test score data and survey responses. This
analysis involved six stages.
Stage 1: Quantitative Analysis of Quantitative Data
The first stage involved the use of descriptive statistics (i.e., descriptive stage; data reduction)
to compute reading comprehension and reading vocabulary scores and compare them to the
normative data. The screenshots for obtaining the descriptive statistics and output are
displayed in Figures 2-4. A series of independent samples t tests (not shown) revealed that the
current sample of doctoral students had statistically significantly higher scores on the reading
comprehension (t = 6.84, p < .0001; effect size = 0.49) and reading vocabulary (t = 11.21, p
< .0001; effect size = 0.80) components of the NDRT than did Brown et al.’s (1993)
normative sample of 5,000 undergraduate students from 38 institutions. However,
disturbingly, approximately 10% of doctoral students attained reading comprehension and
reading vocabulary scores that represented the lower percentiles of this normative sample.
Stage 2: Qualitative Analysis of Qualitative Data
In the second stage, the doctoral students’ perceptions of barriers that prevented them from
reading empirical articles were subjected to a thematic analysis (i.e., exploratory stage; data
reduction) using constant comparison analysis (Glaser & Strauss, 1967). This analysis
6
www.macrothink.org/ije
International Journal of Education
ISSN 1948-5476
2011, Vol. 3, No. 1: E13
revealed the following eight themes that represented students’ perceived barriers to reading
empirical literature: time, research/statistics knowledge, interest/relevance, text coherence,
vocabulary, prior knowledge, reader attributes, and volume of reading.
Stage 3: Quantitative Analysis of Qualitative Data
The themes then were quantitized (i.e., data transformation) such that if a doctoral student
listed a characteristic that was eventually unitized under a particular theme, then a score of
“1” was assigned to the theme for the student response; otherwise, a score of “0” was
assigned. This dichotomization led to the formation of what Onwuegbuzie (2003) called an
inter-respondent matrix of themes (i.e., participant x theme matrix) that consisted only of 0s
and 1s. This inter-respondent matrix of 0s and 1s was entered into the SPSS database,
alongside the other variables. Figure 5 displays part of these data.
The inter-respondent matrix was used to calculate the frequency (i.e., prevalence rate) of each
theme. The steps for conducting the frequency analysis are displayed in Figures 6-8, and the
effect sizes pertaining to three of the themes extracted from qualitative data are presented in
Figure 9.
Stage 4: Quantitative Analysis of Qualitative Data
The fourth stage of Benge et al.’s (2010) SMA involved a principal component analysis to
ascertain the underlying structure of seven of the eight emergent themes (i.e., ...
Purchase answer to see full
attachment
Uncategorized