1 MMSU-CHS-Department of Nursing

LESSON 2a: DESIGNING AND PLANNING PHASE: DETERMINING THE APPROPRIATE

RESEARCH DESIGN

INTRODUCTION OF THE LESSON AND LEARNING OUTCOMES

In Chapter 1, you have learned that research is characterized by a systematic approach
to gathering information to answer questions, which is in contrast to those approaches that
use intuition, seek expert advice, or follow tradition. The research design is the plan for that
systematic approach, conducted in a way that ensures the answer(s) found will be as
meaningful and accurate as possible. The design identifies how subjects will be recruited and
incorporated into a study; what will happen during the study, including timing of any
treatments and measures; and when the study will end.
At the end of this lesson, you must have:
1. discussed the importance of research design;
2. differentiated quantitative and qualitative research designs; and
3. discussed the methods to aggregate research as evidence for nursing
practice

WARM-UP ACTIVITY
Based on your previous learning experiences, make a list of the different research
designs that you know. When finished, kindly take a photo of your answer and post it in this
lesson’s discussion forum.

LEARNING INPUTS
A research design is the overall plan for acquiring new knowledge or confirming
existing knowledge. It is a blueprint for conducting a study. Just as the blueprint for a house
must be individualized to the specific house being built, so much the design be made specific
to a study. A research design is selected with two broad purposes; (1) to plan an approach
that will best answer the research question and (2) to ensure the rigor and validity of the
results.
Elements central to the study design include the presence or absence of a treatment,
number of groups in the sample, number and timing of measurements to be performed,
method of sampling, time frame for data collection, planned comparisons, and control of
extraneous variables.
The purpose of design is to maximize the possibility of obtaining valid answers to
research questions or hypotheses. A good design provides the subjects, the setting, and the
protocol within which these comparisons can be clearly examined.
2 MMSU-CHS-Department of Nursing

Answering the Research Question

Two important considerations are important: (1) the fit of the design to the
research question and (2) the functionality of the design for the purpose of the study.
Fit refers to how well the design matches the question of interest. A research design
must fit the type of question asked in order to provide appropriate and effective
answers. The fit of a design to a research question depends on the function of the
design and on how much is known about the topic of the study. In other words,
different research designs serve different functions and, therefore, are particularly
well suited to one type of research question, but not to another. The functions include
designs for describing or understanding; designs for connecting or relating; and
designs for manipulation and prediction.
Ensuring Rigor and Vitality
Rigor is a strict process of data collection and analysis as well as in term that
reflects the overall quality of that process in a qualitative research. Designs in
qualitative research are more flexible and often are described as “emerging” to
indicate that the design may be altered as the study progresses.
Vitality is used in research to refer both to specific ways that measures can
correctly and accurately reflect their intended variable and to the accuracy of the
overall results. Validity of a study always refer to accuracy or correctness.
Internal validity is the extent to which we can be sure of the accuracy and
correctness of the findings of the study. Thus, it refers to how accurate the results are
within the study itself, or internally. External validity is the extent to which the results
of a study can be applied to other groups or situations. In other words, external validity
refers to how accurately the study provides knowledge that can be applied outside of,
external to, the study. Generalizability is a big aspect of external validity because it
refers to the ability to infer that findings for a particular sample can be applied to the
entire population. It also includes the extent to which the findings from a study in one
setting can be applied to other similar settings. Logically, if a study lacks internal
validity, it automatically lacks external validity.

QUANTITATIVE RESEARCH DESIGNS

Time
Retrospective designs are those in which data are collected about past events
or factors. A retrospective study collect data from patients at some point after they
visit a clinic, perhaps 1 or 2 weeks later, and ask them to recall their level of satisfaction
during their visit. Data are collected at onetime point looking at past experience.
Prospective designs are those in which data are collected about events or
variables as they occur.
3 MMSU-CHS-Department of Nursing

Point of time of measurement

Cross sectional if they collect all data at one point in time. A study of patient
satisfaction that surveys patients as they leave a clinic also is a cross sectional because
data are collected at one-time point for each subject.
Longitudinal if they collect data at different time points. A prospective
longitudinal patient satisfaction study might collect data before a visit to a clinic,
immediately after the visit, and 1 week later, looking for changes in selected variables
over time.

Control

The areas within which research designs seek to create or to impose control
include the sampling and measurement processes.
Control in the sampling process can be imposed by establishing criteria for
inclusion or exclusion. Another method of control is the use of random sampling, by
which the entire population has an equal chance of being asked to be in the study. A
third method is random assignment because all subjects have an equal chance of
being included in any particular group in the study.
Control within the data collection process can be imposed by ensuring the
validity and reliability of the measures or by ensuring that the measurement process
itself is consistent, avoiding instrumentation threats.
Control in measurement also can be imposed by creating comparison group(s)
so that either exposure to the factors studied is manipulated in a controlled fashion.
A design using a comparison group attempts to ensure that two groups are as similar
as possible on most factors that would affect the dependent variable of interest and
assume that they differ clearly in an independent variable.
Study designs that include a control group create a greater level of control by
manipulating the independent variable of interest so that the control group is not
exposed to it, whereas the experimental group is. Again, a dependent variable is
examined for differences to see if the factor manipulated affects that dependent
variable.

Descriptive Designs and Correlational Studies

Descriptive designs function to portray some phenomenon of interest as
accurately as possible. Clearly, it is used to answer research questions that seek to
describe.
Typical descriptive design is used to examine characteristics of a single
sample such as identifying a phenomenon of interest, identifying the variables
4 MMSU-CHS-Department of Nursing

within the phenomenon, developing conceptual and operational definitions of

the variables, and describing the variables.
Comparative descriptive design is used to describe variables and to
examine differences in variables in two or more groups that occur naturally in
a setting
Descriptive case study design examines a single unit, maybe a person,
a family, a nursing unit, or an organization within the context of its real-life
environment.
Correlational studies use a descriptive design to describe interrelationships
among variables. It is used to answer research questions that seek to link or connect.
Both types focus on exerting control through the quality of the measurement-that is,
by using reliable and valid measures and through sampling criteria. These types of
research design may impose control by establishing criteria for inclusion or exclusion
from the study. Remember, this also can be called purposive sampling or use of a
convenience sample.
Descriptive correlational design is use to describe variables and
examine relationships among these variables. Using this design facilitates the
identification of many interrelationships in a situation.
Predictive correlational design on the other hand, is use to predict the
value of one variable based on values obtained for another variables(s). A
predictive study attempts to predict the level of the dependent variable from
the independent variable.
Model testing design are designed specifically to test the accuracy of a
hypothesized causal model (middle range theory). It requires that all variables
relevant to the model be measured.
Descriptive and correlational designs can be longitudinal or cross sectional,
and they can be retrospective or prospective. Decisions about how time is a factor are
based on the nature of the question, the potential sample, and the measures. Some
phenomena, such as growth or productivity, clearly entail a time element that would
make it logical for a researcher to use longitudinal design.

Quasi-Experimental and Experimental Research Design

Quasi-experimental and experimental research designs function to answer
questions involving prediction and the effects of manipulation.
Quai-experimental design facilitates the search for knowledge and
examination of causality in situations in which complete control is not possible. This
was developed to control as many threats to validity as possible in a situation in which
some of the components of true experimental designs are lacking. Quasi-experimental
differs from experimental primarily in the amount of control imposed. Both types
5 MMSU-CHS-Department of Nursing

include control of an independent variable, but a true experimental design always

includes a control group and random assignment to groups. Remember, a control
group is a group of subjects who do not receive an intervention so that the control
group can be compared with those who do receive the intervention.
Experimental design provides the greatest amount of control possible in order
to examine causality more closely. Some examples are:
Pretest-Posttest Design, the most common experimental design used
in nursing studies with experimental and control groups. Multiple groups (both
experimental and control) can be used to great advantage in this design.
Randomized Clinical trial is a carefully designed experimental study
that uses large numbers of subjects to test the effects of a treatment and
compare the results with those of a control group that has not received the
treatment. Subjects are drawn from a reference population, using a clearly
defined criteria, and then are randomly assigned to treatment or control
groups.
Although most quasi-experimental and experimental designs are longitudinal,
it is possible for an experiment to be implemented at only one-time point. Because
experimental designs always involve manipulation of an independent variable, they
are never retrospective. A quasi-experimental design lacks either a control group or
random assignment. It may not include two groups at all.

QUALITATIVE RESEARCH DESIGNS

The goal of qualitative research is to gain knowledge that informs our practice broadly
and holistically, understanding that all knowing is evolving and contextual. The methods focus
on acquiring the richest possible data-that is, data with the greatest complexity and variety.
Therefore, the designs intentionally seek to avoid external control over setting and factors.
Qualitative studies are most often done when we know the least about the topic of interest.
There are three broad functions of qualitative research designs. Designs that function
to facilitate understanding answer descriptive design questions. Design that seek to promote
participation or immersion answer both description and connection. Designs that seek to link
ideas and concepts answer questions of connection or relationship.
Phenomenology
Phenomenology, or the phenomenological method, is used to discover and
develop understanding of experiences as perceived by those living the experience.
There are several variations on the phenomenologic method, but in general, the
method includes identifying the people who are living or have lived the experience of
interest and seeking, usually through unstructured interviewing, their perceptions. As
data are collected, the researcher uses the processes of intuiting, analyzing, and
describing to discover essential themes in the experience of the phenomenon.
6 MMSU-CHS-Department of Nursing

Phenomenology uses a spiraling process of data collection and analysis, and

detailed filed notes of observations during data collection augment the richness and
fullness of data. Data are collected until all information is redundant of previously
collected data –until saturation occurs. Sampling in phenomenology is always
convenience sample. The ability to communicate about and lived the experience is the
only criteria on who can be a participant.
Example: “Needs of family members of patients with traumatic brain injury”

Ethnography
Ethnographic method originated in the discipline of anthropology, and its
purpose is for the researcher to immerse himself or herself in a culture to describe a
phenomenon or phenomena within the context of that culture. A closely related
method that was developed by Leininger within nursing is called ethonursing. Culture
exists, even though it is not visible, and the only way to know a culture is to get both
an insider’s view – emic perspective and an outsider’s view – etic perspective.
Controlling the environment or aspects of the study is not part of this method.
The researcher tries to become part of the culture studied to acquire an insider’s
understanding so that he or she can then translate it into a common language
understood by those outside the culture. Data are collected as they happen and as
opportunities present themselves. Therefore, there is no structured format for the
collection of data.
Example:” Health beliefs and practices of the Malawa tribe”

Grounded Theory
Grounded theory method is to study interactions to understand and recognize
links between ideas and concepts or, in other words, to develop theory. This method
is best used to study social processes and structures hence the focus on links and
interactions among ideas or categories.
Grounded theory methods often incorporate time into the study because the
focus usually is on processes or change. Sampling usually will be purposive-that is,
purposely seeking participants experiencing the process or changes under study. Data
collection can include interviews and careful observations of interactions and
processes.

Example: “Feeling safe: The psychosocial needs of ICU patients”

Historical
Historical research method is used to answer questions about links in the past
to understand or to plan the future. This method requires the researcher to define a
7 MMSU-CHS-Department of Nursing

phenomenon in a manner that can be clearly delineated so that data sources can be
identified.
Data sources may include records, videotapes, photographs, and interviews
with people involved in the phenomenon or review of published reports. A focus of
data collection includes evaluation of data sources for their reliability.
Example: “Historical method of side rail use in American hospitals”

Threats to the Rigor of a Qualitative Study

The rigor of a qualitative study is reflected in the consistency, trustworthiness,
transferability, and credibility of the study. Qualitative designs are based on distinct
philosophical perspectives and have clearly defined systematic methods associated
with each design. When considering these concepts in reference to the overall rigor
of a study, we must consider consistency in the application of the study design
throughout the study, or consider credibility given the researchers’ accurate use of
the study method.
Threats to Internal Validity
Threats to internal validity are potential problems that can affect the accuracy
or correctness of findings within a study. These include the following:
History is some factor outside those examined in a study affecting the
outcome or dependent variable.
Maturation refers to change in the dependent variable simply because
of the passage of time.
Testing refers to changes in a dependent variable that that result
because it is being measured or because of the measure itself.
Instrumentation refers to changing the measures used in a study from
one time to another.
Mortality refers to loss of subjects from a study because of consistent
factor related to the dependent variable. Occasionally, the loss of subjects is
from death. At other times, mortality refers to subjects withdrawing from a
study.
Selection bias refers to subjects having unique characteristics that in
some manner relate to the dependent variable.
In summary, internal validity is threatened because some outside factor
(history) or time (maturation) affects the dependent variable, because the
measurement process itself (testing) or changes in a manner (instrumentation)
affects results for the dependent variable, or because the sampling process is biased
by loss of subjects (mortality) or selection of subjects (selection bias).
8 MMSU-CHS-Department of Nursing

Threats to External Validity

Threats to external validity are potential problems in a study that affect the
accuracy of the results for samples and settings other than those of the study itself.
They include the following:
Reactivity effects/Hawthorne effect refer to the responses of subjects
to be studied. Reactivity can occur in a broader sense simply because subjects
know that they are being studied.
Measurement effects are changes in a results of a study resulting from
various data collection problems.
Novelty effects occur when the knowledge that what is being done is
new and under study somehow affects the outcome, either favorably or
unfavorably.
Experimenter effects occur when some characteristic of the researcher
or data collector influences the study results.

A Contrast of Quantitative and Qualitative Characteristics

Element Quantitative Qualitative

View of reality Reality is objective and can Reality is constructed by the
be seen and measured individual
View of time Reality is relatively constant Reality is continuously
constructed
Context Reality can be separated Reality is embedded in its
from context context
Researcher approach Objective, detached Personally involved
Populations studied Samples that represent Individual cases,
overall populations as represented as informants
subjects
Measures Human behavior or other Study the meanings that
observable phenomena individuals create
Observations Analyze reality as definable Make holistic observations
variables of the total context
Design Preconceived and highly Emergent and fluid,
controlled adaptable to informants’
view
Analysis Descriptive and inferential Analytic induction to
statistics determine meaning
Generalization Use inference to generalize Transfer knowledge from
from a sample to a defined case analysis to similar cases
population
Reports Objective, impersonal Interpretive reports that
reports in which the reflect the researcher’s
9 MMSU-CHS-Department of Nursing

researcher’s opinions are reconstruction of the

undetectable meaning of the data

FINDING AND AGGREGATING EVIDENCE

Many methods exist for the identification and aggregation of research for translation
into practice, including systematic reviews, meta-analysis and meta-synthesis. A discussion of
what the mixed method is all about will also be done. The following information will elucidate
the meanings of these terms.

THE SYSTEMATIC REVIEW

A highly structured and controlled search of the available literature that minimizes the
potential for bias and produces a practice recommendation as an outcome. Systematic
reviews can focus on patient concerns, the prevalence of problems, or the effectiveness of
healthcare interventions. These reviews are critical for evidence-based practice because they
summarize the numerous and sometimes contradictory findings in an unbiased, methodical
way.
The Cochrane Library (www.cochrane.org) is widely recognized as one of the most
useful sources of high-quality reviews of healthcare interventions; because of their through
searches and their methodological rigor on quantitative analysis and experimental designs.
As the name implies, planning the search strategy for a systematic review involves a
sequence of carefully considered questions:
1, Determine the background for the review: Why is it important to have a systematic
review for this intervention? How was the need identified?
2. State the main review question: What is the goal of the review? Is the review to test
a specific existing intervention or to come up with “best practices”?

3. Develop inclusion and exclusion criteria: Who are the patients of interest? What age
groups, diagnoses, or other conditions are of interest? Which specific patient groups will not
be included? Which interventions and outcomes are of interest?
4. Devise a search strategy: What are the sources of studies that will be searched
(including published, unpublished, and “gray literature” such as conference proceedings)?
What are key journals that must be hand-searched?
5. Develop study selection criteria: What search terms will be used? What types of
studies are acceptable (for example, experimental, descriptive, qualitative)? What time frame
will be considered?
6. Determine study quality criteria: What quality indicators will be used to appraise
articles for inclusion? What quality level is acceptable overall? What quality problems warrant
exclusion from the study?
10 MMSU-CHS-Department of Nursing

META-ANALYSIS
A statistical method of aggregating the results of quantitative studies so an overall size
can be evaluated. When experiments are replicated on similar populations using standard
measures, it is possible to sum up the aggregate impact of the intervention on an outcome.
The strength of a comprehensive meta-analysis is its ability to reveal structural flaws and
sources of bias across primary research studies. When the results of multiple studies are
compared, methodological weaknesses become apparent and true effects are revealed.
The challenge of meta-analysis is to find a sufficient number of studies that used
similar populations, measures and statistics. Meta -analysis is particularly challenged to
quantify the size of a common treatment effect because of the clinical diversity of the trials
and the potential differences among patients in the trials. Available software programs can
analyze disparate types of data for commonalities.

QUALITATIVE META-SYNTHESIS
Qualitative meta-synthesis is the development of overarching themes about the
meaning of human events based on a synthesis of multiple qualitative studies. Meta-synthesis
represents a family of methodological approaches to developing new knowledge based on
rigorous analysis of existing qualitative research findings. The focus of metas-ynthesis is on
recurrent themes.

MIXED- METHODS or MULTIMETHOD RESEARCH

A research approach that combines quantitative and qualitative elements; it involves
the description of the measurable state of phenomenon and the individual’s subjective
response to it.
Advantages include: (1) the two methods have complementary strengths and
weaknesses; (2) an integrated approach can lead to theoretical and substantive insights into
the multidimensional nature of reality; (3) multimethod research can provide feedback loops
that augment the incremental gains in knowledge from a single-method study; (4)
confirmation of hypotheses through multiple types of data can strengthen study validity; an
(5) if findings are inconsistent, a careful scrutiny of the discrepancies could push the line of
inquiry further.
Despite the advantages of multimethod research, there are several potential
obstacles. These include epistemologic biases, high costs, inadequate researcher training, and
publication biases.
Mixed methods are commonly used in descriptive studies. Mixed methods may be
used to describe both the measurable state of a phenomenon and the individual responses
to it. For example, a mixed method may be used to describe the rate of hand washing on a
nursing unit (quantitative) as well as the nurses’ perceptions about the importance of hand
washing (qualitative).
11 MMSU-CHS-Department of Nursing

CENTRAL ACTIVITIES
Activity 1
Identify if the following elements describe quantitative or qualitative research by
checking the appropriate column. Once finished, take a photo and post it in the discussion
forum created for this lesson.

Element Quantitative Qualitative

Soft science
One reality, focus is concise
and narrow
Tests theory
Organismic; whole is greater
than the sum of its parts
Subjects
Researcher is part of the
process
Measurable
Literature review as the
study progresses
Basic element of analysis is
numbers
Context free

Wrap-up Activity
Look for two researches online. Identify the research design and note down the
research questions/objectives used. Follow the format below. Once finished, take a photo
and post it in the discussion forum created for this lesson.

Title of the Research Research Design Used Research

questions/objectives

Assessment
A summative assessment will be administered to you. Be sure to get at least 60% of
the test items.
12 MMSU-CHS-Department of Nursing

REFERENCES
Barcelo, T. I. (2013). Burns and Grove’s understanding nursing research: Building an evidence-
based practice (Philippine edition). Elsevier (Singapore) Pte Ltd.
Houser, J. (2018).Nursing research: Reading, using and creating evidence (4th ed.). Jones &
Bartlett Learning. MA, USA
Macnee, C.L. & McCabe, S. (2008) Understanding nursing research: reading and using
research in evidence-based practice (2nd ed.). Wolters Kluwer.
Polit, D.F., & Beck, C.T. (2010). Essentials of nursing research: Appraising evidence for
nursing practice (7th ed.). Lippincott Williams &Wilkins.