Applied Nursing Research 31 (2016) 1–5

Contents lists available at ScienceDirect

Applied Nursing Research

journal homepage: www.elsevier.com/locate/apnr

Guidance for using mixed methods design in nursing practice research☆

Lenny Chiang-Hanisko, PhD, RN ⁎, David Newman, PhD, MA, Susan Dyess, PhD, RN,
Duangporn Piyakong, PhD, RN, Patricia Liehr, PhD, RN
Florida Atlantic University, Christine E. Lynn College of Nursing, Boca Raton, FL, 33431

a r t i c l e i n f o a b s t r a c t

Article history: The mixed methods approach purposefully combines both quantitative and qualitative techniques, enabling a
Received 18 May 2015 multi-faceted understanding of nursing phenomena. The purpose of this article is to introduce three mixed
Revised 17 November 2015 methods designs (parallel; sequential; conversion) and highlight interpretive processes that occur with the syn-
Accepted 2 December 2015
thesis of qualitative and quantitative findings. Real world examples of research studies conducted by the authors
Available online xxxx
will demonstrate the processes leading to the merger of data. The examples include: research questions; data col-
Keywords:
lection procedures and analysis with a focus on synthesizing findings. Based on experience with mixed methods
Mixed methods design studied, the authors introduce two synthesis patterns (complementary; contrasting), considering application for
Nursing practice research practice and implications for research.
© 2015 Elsevier Inc. All rights reserved.

Practicing nurses recognize the complexity of the human condition
and the challenge of understanding health circumstances for unique in-
dividuals. Mixed methods appeals to nurse researchers because it offers
an approach for broadening understanding that captures multi-faceted
dimensions of health circumstances. Mixing methods is a process of
combining both quantitative and qualitative techniques. According to
Sandelowski (2014), it is “a new way of recognizing and speaking
about the methodological and design mixes constituting all empirical
inquiry” (pp. 6–7).
Even though mixed methods is gaining in popularity (Creswell,
Klassen, Plano Clark, & Clegg Smith, 2011; Pluye & Hong, 2014), there
is still much debate as to when mixing methods is appropriate and
how meaningful mixing occurs. Mixed methods, like other systematic
research approaches, includes: consistency between the research ques-
tion, purpose, and methodological choices; verifiable and transparent
techniques that demonstrate trustworthiness; potential for replicabili-
ty; opportunity for self-correction; and ability to explain the phenome-
na under investigation (Newman & Hitchcock, 2013). However, the
“when, why, and how” of mixing persist as relevant questions. The
question of “when” is easy. Like all well-designed research studies, the
methods must fit the question of interest. The types of questions that
are most appropriate for mixed methods designs are ones for which nei-
ther qualitative nor quantitative approaches alone could adequately an-
swer the question (Creswell et al., 2011; Newman, Newman, &
Newman, 2011). The question of “why” is also easy. That is, to provide
☆ Conflict of Interest Statement: The authors declare no conflicts of interest.
⁎ Corresponding author. Tel.: +1 561 297 2937; fax: +1 561 297 2416.
E-mail addresses: lchiangh@fau.edu (L. Chiang-Hanisko), Dnewma14@fau.edu
(D. Newman), sdyess@fau.edu (S. Dyess), dpiyakon@my.fau.edu (D. Piyakong),
pliehr@fau.edu (P. Liehr).
insight from multiple angles (Teddlie & Tashakkori, 2009). The question
of “how” methods are merged tends to produce the greatest difficulty. Im-
plicit in “how” methods are mixed is a question of “where” the mixing oc-
curs, whether at the philosophical, data collection/analysis or interpretive
phases of the study. Guest (2013) has introduced the idea of “points of in-
terface” that could occur at the data collection or analysis phase or at the
interpretive phase when qualitative and quantitative results (inferences)
are generated or when interpretive efforts are directed to merge qualita-
tive and quantitative results (meta-inference).
The primary goal of this paper is to introduce three mixed methods
designs (parallel; sequential; conversion) and focus on the point of in-
terface occurring at the meta-inference, where interpretative processes
demand synthesis of qualitative and quantitative inferences. Three real
world examples will illustrate each mixed methods design. The exam-
ples are based on three separate studies undertaken by members of
this authorship team. Each of the studies has contributed to the ad-
vancement of a program of research with older adults. The steps in
each example provide essential information about study details but
the focus is on “how” qualitative and quantitative findings can come to-
gether at the interpretive point of interface to guide nursing practice
and research.
1. Parallel mixed methods
In the parallel mixed methods approach, data collection, analysis
and inference generation occur side-by-side to address distinct research
questions (Tashakkori & Newman, 2010). At least two inferences, one
qualitative and one quantitative, are reported. Then, these inferences
are synthesized at the interpretive point of interface. In the first exam-
ple the parallel design was selected to explore health in ethnically di-
verse older adults living in the community with chronic illness.

http://dx.doi.org/10.1016/j.apnr.2015.12.006
0897-1897/© 2015 Elsevier Inc. All rights reserved.
2 L. Chiang-Hanisko et al. / Applied Nursing Research 31 (2016) 1–5
1.1. Research questions
Two research questions were posed: (1) What is the relationship be-
tween ethnicity and perception of health for ethnically diverse, communi-
ty dwelling older adults (quantitative)? (2) What strategies do ethnically
diverse older adults use to manage health challenges (qualitative)?
1.2. Data collection, analysis, and inference findings
The quantitative data were acquired from an existing database with-
in a Healthy Aging Registry of 350 community-dwelling older adults
equally representing four diverse ethnic groups (African American
[AA], Afro-Caribbean American [AC], European-American [EA] and
Hispanic-American [HA]). The researcher accessed participant scores
on the SF-36 (Ware, 1993), a measure that assesses self-perception of
physical and mental health. Cronbach's alpha exceeds .70 for all sub-
scales of the SF-36 in testing with adults, and there is substantial support
for the content, criterion, construct and predictive validity (Tsai, Bayliss, &
Ware, 1997). Differences in physical and mental health perception across
ethnic groups were examined using analysis of variance (ANOVA).
Qualitative data were obtained from the same existing Healthy
Aging Registry, using a stratified random selection of adults who
consented to be contacted for future research. A total of 16 participants
were selected; the adults were evenly distributed from each ethnic
group, spoke English and were living with chronic illness. A theory-
guided story-gathering approach (Smith & Liehr, 2014) was used to
query strategies for managing health challenges. Interviews were con-
ducted until saturation was reached; they were audio-recorded and
transcribed verbatim. Content analysis (Hsieh & Shannon, 2005) was
used to address strategies for managing health challenges.
1.3. Quantitative findings
On average, EA subjects perceived themselves to be more physically
healthy than AA or AC subjects as they scored 2.9 points higher than AA
subjects (p = .04) and 4.3 points higher than AC subjects (p = .003) on
the physical health scale of the SF-36. From the perspective of mental
health, AA subjects perceived greater mental health than EA or HA sub-
jects as they scored 4 points higher than EA subjects (p = .007) and 9
points higher than HA subjects (p b .001) on the mental health scale
of the SF-36. The inference drawn from quantitative findings is that:
EA subjects perceive greater physical health than AC and AA subjects,
but not HA subjects; and, AA subjects perceive greater mental health
when compared to EA and HA subjects, but not AC subjects.
1.4. Qualitative findings
All participants, regardless of ethnicity, identified “life adjustment”
as essential to managing wide-ranging health challenges associated
with living in the community with a chronic illness. All ethnic groups
engaged in individually unique approaches to “manage the best they
can” with specific patterns such as: relying on others, trusting in
healthcare, and hoping for the best. Spiritual activity was identified in
all but the EA group as a viable approach for managing health chal-
lenges. These results constitute the inferences emerging from the qual-
itative data analysis.
1.5. Interpretive point of interface: synthesis of qualitative and
quantitative data
The first synthesized finding is that there are significant ethnic differ-
ences in mental and physical health (quantitative) in spite of older adults'
consensus view that managing the challenges of chronic illness demands
life adjustment (qualitative). The second finding is that AA and AC ethnic
groups are most similar. The most similarity is that both groups had
higher mental health scores than other ethnic groups (quantitative). In
addition, both groups clearly identified that their spiritual perspectives
contributed to managing health challenges (qualitative).
The synthesized findings at the interpretive point of interface have
implications for clinical nursing practice with older adults, but targeted
future study is warranted. Still, the findings emphasize the importance
of culturally unique care, particularly related to spiritual activity as a vi-
able resource for promotion of mental health in Afro-American and
Afro-Caribbean American older adults. In this example, it is as though the
qualitative and quantitative data are two sides of the same coin. Practicing
nurses implicitly recognize this holistic perspective, knowing that numer-
ical data are infused with stories that promise enhanced understanding.
The parallel mixed methods approach, when applied with thoughtful in-
terpretation at the intersection of qualitative and quantitative data, has
the potential to bring structure to capturing a holistic perspective.
2. Sequential mixed methods
The sequential mixed method approach takes place when the qual-
itative and quantitative methods occur in two separate time-ordered
phases, and the collection and analysis of one type of data follows and
is dependent on the collection and analysis of the other type. The design
can be sequential exploratory (qualitative followed by quantitative) or
sequential explanatory (quantitative followed by qualitative). This ex-
ample presents the sequential exploratory design based upon the use
of Q-methodology (Brown, 1996; Newman & Ramlo, 2010). This design
was selected because the researcher wanted to use qualitative findings
to inform development of typologies that were relevant for a larger pop-
ulation (Creswell & Plano Clark, 2011). A sequential explanatory design
may be used when the finding of a quantitative study could be further
explained and interpreted by using a qualitative method.
2.1. Research questions
Two research questions are used to exemplify the sequential ap-
proach: (1) How do ethnically diverse older adults describe perceptions,
attitudes, and approaches for pain management associated with
polypharmacy (qualitative)? (2) What are the polypharmacy typologies
descriptive of ethnically diverse older adults (quantitative)?
2.2. Data collection, analysis, and inference findings
Using the sequential exploratory design, the first step was to collect
and analyze the qualitative data. Twenty face-to-face interviews were
conducted to obtain detailed information about participants' experience
with medication practices that could lead to problems with
polypharmacy. Qualitative data analysis began after interviews were tran-
scribed and then validated using trustworthiness estimates such as mem-
ber checking. Using thematic analysis, each meaningful statement in the
interview was assigned a code that captured its meaning. The codes
were grouped into categories and then clustered into themes that ad-
dressed perceptions, attitudes, and actions regarding medication usage
among the participants.
2.3. Qualitative findings
Thirty four statements derived from the qualitative data analysis cre-
ated items to be used in the next phase of analysis. Examples of items in-
clude “taking too much pain medicine is harmful to my health”
(perception); “it is better to hang in there and tough it out without the
pain medication” (attitude); and “I will take my pain medication before
I start to hurt because I don't want to feel any pain” (action). In its final
form, the 34 items are representative and summative of participant re-
sponses to the qualitative research question. These items are the inference
for the qualitative phase, leading to the quantitative strand of the study.
 L. Chiang-Hanisko et al. / Applied Nursing Research 31 (2016) 1–5
2.4. Transition to the quantitative study phase
Since this study was designed based on Q methodology, the 34 items
were numbered randomly and typed onto cards for use in a sorting pro-
cedure that is essential to the analysis process (Brown, 1996). The sorting
procedure required participants to read each of the 34 statements and
then rank-order each statement on a continuum from “strongly agree”
to “strongly disagree”. The rank-ordered scores noting agreement/dis-
agreement were then analyzed using a Q-factor analysis. In this phase of
integrating qualitative and quantitative data, statistical analysis enabled
clustering of the 34 items into factors based upon similarities.
2.5. Quantitative findings
The outcome of factor analysis was identification of typologies that
could be used to predict polypharmacy. A “self-management centered”
typology included participants who took individual responsibility for
managing their pain medication rather than relying on physician's recom-
mendations. A “physical functioning centered” typology included partici-
pants with lower physical function, who had multiple chronic illnesses
and would take multiple medications to continue their daily activities.
2.6. Interpretive point of interface: synthesis of qualitative and
quantitative data
The interpretive point of interface that merged qualitative and quan-
titative data in this example has subtly occurred as the inferences from
the qualitative strand of the study (34 items) became the substance of
quantitative data collection (sorting from “strongly agree” to “strongly
disagree”) and analysis (Q-factor analysis). The synthesized findings of
two distinct typologies, “self-management centered” and “physical
functioning centered,” are important to understanding patterns of med-
ication use for ethnically diverse older adults.
Knowledge of these typologies can alert practicing nurses to pat-
terns that distinguish older adults but also uniquely predispose them
to polypharmacy. Older adults who are “self-management centered”
may freely add pain medicines without consulting or informing their
health care provider; those who are “physical function centered” may
escalate pain medicine doses to enable preferred physical activity with-
out evaluating potential complications. These considerations, taken to
practice, enhance the nurse's ability to detect polypharmacy threats in
older adults.
3. Conversion mixed methods
The conversion mixed methods approach can begin with either
qualitative or quantitative data so that words are converted to numbers
or numbers are converted to words (Tashakkori & Newman, 2010). In
the example for the conversion mixed methods approach the researcher
began with narrative data, conducted qualitative analysis and arrived at
the qualitative inference. Words were then converted into numbers en-
abling statistical analysis; this process is sometimes referred to as
quantitizing (Sandelowski, 2014). Quantitizing enabled generation of
the quantitative inference.
3.1. Research questions
Two research questions are used to exemplify the conversion ap-
proach: (1) What are the turning points described by veterans in stories
about surviving the bombings of Pearl Harbor (qualitative)? (2) What
percentage of personal pronouns, positive/negative emotion words
and cognitive process words are used by Pearl Harbor veterans in
their stories of survival (quantitative)?
3
3.2. Data collection, analysis and inference findings
When qualitative data are converted to quantitative data, the re-
searcher can collect written or oral narrative data. In this example, a
theory-based oral format, the story-path, was used to gather data
(Smith & Liehr, 2014). Story-path guides the researcher to engage partic-
ipants in their current, past and hoped-for future experience, in this case
the experience of getting along day-to-day following the trauma of war-
time bombing (Liehr, Nishimura, Ito, Wands, & Takahashi, 2011). The
stories of 23 Pearl Harbor survivors were audio-recorded, transcribed
and analyzed. To address the qualitative research question, the transcrip-
tions were analyzed for turning points that occurred since surviving the
bombings of Pearl Harbor. Turning points were defined as “shifts in
ones way of being ‘day to day’ over time” (Liehr et al., 2011, p. 217).
3.3. Qualitative findings
The turning points had been previously analyzed by a team of inter-
national researchers who found that surviving was characterized by:
(1) coming to grips with and responding to the reality of a Japanese at-
tack; (2) honoring war memories but setting them aside to get on day-
by-day; and (3) embracing connection as a source of comfort and un-
derstanding (Liehr et al., 2011). These three turning points constitute
the inference emerging from the qualitative analysis.
3.4. Transition to quantitative study phase
For the quantitative analysis, transcripts were prepared for analysis
with Linguistic Inquiry and Word Count (LIWC) software developed
and tested by Pennebaker and colleagues (Pennebaker, Booth, & Francis,
2007; Pennebaker, Chung, Ireland, Gonzales, & Booth, 2007; Pennebaker
& King, 1999). The LIWC program is a word-based computerized text
analysis software, which discerns linguistic categories, including those
selected for this analysis: negative (eg: tense, upset, turmoil) and posi-
tive (eg: calm, relaxed, happy) emotion word-use (alpha reliabilities
from .71 to .97); cognitive process (eg: think, understand; analyze)
word-use (alpha reliabilities from .76 to .97); and personal pronoun-
use, such as “I” and “we” (alpha = .74 and .70 respectively). Word-
use for each category is presented as a percentage of words in the spe-
cific category compared to the entire transcript. Descriptive statistics
were used to analyze the data.
3.5. Quantitative findings
There were more than triple the percentage of “I” (X = 7.1 ± 4.5)
compared to “we” words (X = 2.1 ± 2.5) and comparable percentages
of positive (X = 1.4 ± 1.7) and negative (X = 1.0 ± 1.3) emotion
words. Cognitive process words comprised an average of 6.5% (sd =
3.6) of the words used by the Pearl Harbor survivors. The inference
from this quantitative analysis indicates that seven decades after
experiencing wartime bombing, Pearl Harbor survivors focused on cog-
nitive processes rather than emotions and stories were shared from the
“I” rather than “we” perspective.
3.6. Interpretive point of interface: synthesis of quantitative and
qualitative data
The synthesized finding which specifically merges qualitative and
quantitative inferences focuses on human connection. In spite of en-
dorsing human connection as a meaningful turning point (qualitative),
these older adults expressed themselves with triple the number of “I”
words compared to “we” words (quantitative). This finding at the inter-
pretive point of interface warrants further study. For instance, a ques-
tion about the enabling power of owning one's story of trauma could
be considered as it relates to human connection for older adults.
4 L. Chiang-Hanisko et al. / Applied Nursing Research 31 (2016) 1–5
Converting stories to numbers has practical appeal for nurses, partic-
ularly nurse researchers in practice settings. Stories are ubiquitous and
often overlooked “vital signs” shared by patients in everyday nursing
practice. By their very nature, thoughtfully gathered stories can serve
as evidence that informs nursing care. When systematic analysis, inclu-
sive of qualitative and quantitative components is applied, nursing prac-
tice stories are reframed. Stories become evidence with narrative and
numerical impact. Juxtaposed narrative and numerical data have the
potential to highlight consistencies and irregularities providing a foun-
dation for next steps in the direction of nursing practice or research.
4. Discussion
Symonds and Gorard (2010) indicate that mixing truly occurs when,
“elements of the research process are used to construct, transform and
influence each other” (p. 13). Our experience with “true” mixing was
an arduous two-year process that included thoughtful reflection about
findings from studies we were conducting; dialogue with colleagues, in-
cluding doctoral students, who challenged our thinking; and presenta-
tions at national meetings where audience members raised thought
provoking questions that led us to reconsider the interpretive strategies
we were using. The result of this process is reflected in Fig. 1 illustrating
the logical process for mixing methods to generate the meta-inference,
noting the move from meta-inference to nursing practice and research.
It is not intended as a comprehensive description addressing complex
nuances but rather a roadmap for the clinical researcher wishing to em-
bark on study of a nursing practice phenomenon that can best be served
by using a mixed methods design.
This scholarly process took us to the interpretive point of interface
where quantitative and qualitative findings are synthesized, a critical
“mixing” nexus that culminated in one of two patterns of quantitative/
qualitative relationship: complementary or contrasting patterns. When
findings were complementary, the qualitative/quantitative inferences in-
formed each other and were more readily translated for application to
nursing practice. For instance, in the description of the parallel design
for mixed methods, the researcher found that Afro-American and Afro-
Caribbean American people were similar in that both groups perceived
that they had better mental health than the other ethnic groups (quanti-
tative) and both endorsed spiritual perspective as a resource that contrib-
uted to managing their chronic health challenges (qualitative). Nurses are
encouraged to first consider how this finding fits with existing literature
but there is likelihood that these complementary findings could readily
be applied to practice when working with these populations. For instance,
based on these findings, consideration of spiritual perspectives may rise
as a priority for assessment when mental health challenges are suspected
in Afro-American or Afro-Caribbean American patients. It is important to
note that the emergence of practice applications with complementary
findings, does not negate potential for the clinical nurse researcher to
pose a next research question based on the synthesized findings. When
findings are complementary, we are suggesting that practice applications
are in the foreground while research implications are in the background.
While one study is never enough to change practice, the principle of
transferability allows practice consideration for clinical nurses.
Fig. 1. Logical process for mixing methods.
When the quantitative/qualitative inferences present a contrasting
pattern there is a natural tendency to pose the next research question,
constructing a direction for further study. For instance in the description
of the conversion design of mixed methods, the researcher found that
survivors of Pearl Harbor endorsed human connection as an important
turning point in moving along over years (qualitative), yet their use of
the personal pronoun “I” occurred three times more frequently than
“we” in their stories (quantitative). This finding generated curiosity
that may not have immediate application for practice but could readily
indicate a direction for research that would first lead the researcher
to the literature to explore what is known about the relationship be-
tween owning one's story of trauma and human connection. In the
circumstance of contrasting findings, research moves to the foreground
while practice is in the background. What is learned with contrasting
findings may eventually be applied to practice but not without
further exploration.
4.1. Synthesizing quantitative and qualitative findings
Although there are general descriptions for synthesizing quantita-
tive and qualitative findings (creating a meta-inference) (Creswell &
Plano Clark, 2011; Teddlie & Tashakkori, 2009), there is very little de-
tailed guidance. This synthesis is ultimately an interpretive endeavor
that demands time and thought. It brings to mind the idea of
“wallowing while waiting” introduced by Smith (1988) decades ago.
In the case of mixed methods, the researcher wallows with the qualita-
tive and quantitative inferences, immersed in the nuances of findings
until clarity begins to take shape. Wallowing is a unique-to-person ex-
perience. It is a spirit that a researcher brings to the interpretive process
and by its very nature it defies detailed description. In spite of this defi-
ance the authors have attempted to shed some light on the rigor neces-
sary for “true” mixing where quantitative and qualitative findings are
synthesized at the meta-inference point of interface.
5. Conclusion
This paper summarizes and simplifies what we have learned about
mixing methods while working on the studies provided as examples.
The mixed methods design has been described through three approaches,
namely parallel, sequential, and conversion. The choice is a commitment
to embrace the craft of true mixing, knowing that the research dimen-
sions will inform each other, providing guidance for practice and research.
We invite clinical researchers to consider studying complex human phe-
nomenon through the use of mixed methods design.
Acknowledgement
The authors would like to acknowledge Drs. Ruth Tappen and Joseph
Ouslander from Florida Atlantic University for access to the Health
Aging Registry dataset used in this article for the parallel mixed
methods research study.
 L. Chiang-Hanisko et al. / Applied Nursing Research 31 (2016) 1–5
References
Brown, S. R. (1996). Q methodology and qualitative research. Qualitative Health Research,
4, 561–567.
Creswell, J. W., Klassen, A., Plano Clark, V., & Clegg Smith, C. (2011). Best practice for mixed methods
research in the health science. Bethesda, MD: Office of Behavior and Social Sciences Research,
National Institute of Health (Retrieved from http://obssr.od.nih.gov/mixed_methods_
research/pdf/Best_Practices_for_Mixed_Methods_Research_acknowledgement.pdf).
Creswell, J. W., & Plano Clark, V. (2011). Designing and conducting mixed methods research
(2nd ed.). Thousand Oaks, CA: Sage.
Guest, G. (2013). Describing mixed methods research: An alternative to typologies.
Journal of Mixed Methods Research, 7(2), 141–151.
Hsieh, H., & Shannon, S. E. (2005). Three approaches to qualitative content analysis.
Qualitative Health Research, 15(9), 1277–1288.
Liehr, P., Nishimura, C., Ito, M., Wands, L. M., & Takahashi, R. (2011). A lifelong journey of
moving beyond wartime trauma for survivors from Hiroshima and Pearl Harbor.
Advances in Nursing Science, 34(3), 215–228.
Newman, I., & Hitchcock, J. H. (2013). Applying an interactive quantitative-qualitative
framework: How identifying common intent can enhance inquiry. Human Resource
Development Review, 12(1), 36–52.
Newman, I., Newman, D., & Newman, C. (2011). Writing mixed research articles: Method-
ological considerations to help you get published. In T. Rocco, & T. Hatcher (Eds.),
Demystifying the writing and publishing process: A guide for emerging scholars. San
Francisco, CA: Jossey-Bass Press.
Newman, I., & Ramlo, S. (2010). Using Q methodology and Q factor analysis in mixed
methods research. In A. Tashakkori, & C. Teddlie (Eds.), Handbook of mixed methods
in social & behavior research (pp. 505–530) (2nd ed.). Thousand Oaks, CA: Sage.
5
Pennebaker, J. W., Booth, R. J., & Francis, M. E. (2007a). Operator's manual: Linguistic inqui-
ry and world count: LIWC2007. Austin, TX: LIWC.net.
Pennebaker, J. W., Chung, C. K., Ireland, M., Gonzales, A., & Booth, R. J. (2007b). The devel-
opment and psychometric properties of LIWC2007. Austin, TX: LIWC.net.
Pennebaker, J. W., & King, L. A. (1999). Linguistic styles: Language use as an individual dif-
ference. Journal of Personality and Social Psychology, 77(6), 1296–1312.
Pluye, P., & Hong, Q. N. (2014). Combining the power of stories and the power of num-
bers: Mixed methods research and mixed studies reviews. Annual Review of Public
Health, 35, 29–45.
Sandelowski, M. (2014). Unmixing mixed-methods research. Research in Nursing &
Health, 37, 3–8.
Smith, M. J. (1988). Wallowing while waiting. Nursing Science Quarterly, 1(1), 3.
Smith, M. J., & Liehr, P. (2014). Story theory. In M. J. Smith, & P. Liehr (Eds.), Middle range
theory for nursing (pp. 225–252) (3rd ed.). New York: Springer.
Symonds, J., & Gorard, S. (2010). Death of mixed methods? Or the rebirth of research as a
craft. Evaluation & Research in Education, 23(2), 121–136.
Tashakkori, A., & Newman, I. (2010). Mixed methods: Integrating quantitative and qual-
itative approaches to research. In B. McGaw, E. Baker, & P. P. Peterson (Eds.), Interna-
tional encyclopedia of education (pp. 514–520) (3rd ed.). Oxford, UK: Elsevier.
Teddlie, C., & Tashakkori, A. (2009). Foundations of mixed methods research: Integrating
quantitative and qualitative approaches in the social and behavioral sciences. Thousand
Oaks, CA: Sage.
Tsai, C., Bayliss, M. S., & Ware, J. E. (1997). SF-36 health survey annotated bibliography: Sec-
ond edition (1988-1996). Boston, MA: New England Medical Center Health Assess-
ment Lab.
Ware, J. E. (1993). SF-36 health survey: Manual and interpretation guide. Boston, MA: The
Health Institute, New England Medical Center.