This document discusses developing a research question. It states that formulating the right research question is important as it guides the study. A good question is developed through selecting a topic of interest and exploring it in the literature to identify a problem and refine it into a question. Key aspects that shape the question are evaluating its importance and feasibility, specifying the population and variables, and developing a rationale supported by literature. The research question guides the objectives and hypotheses.

1. RESEARCH
QUESTION
PORTNEY, L.G. AND WATKINS, M.P. (2008) FOUNDATIONS OF CLINICAL
RESEARCH: APPLICATIONS TO PRACTICE. 3RD EDITION. PRENTICE HALL.
NEW YORK.

2.  Research is about answering
questions. But before we can get to
the answers, we must be able to ask
the right question. What we ask will
depend on our goal.
 This is the most important and often
most difficult part of the research
process, because it controls the
direction of all subsequent planning
and analysis.

3. THE PROCESS STARTS WITH:
selection of a research topic
that sparks some interest.
exploration of that topic by
examining issues in clinical
practice and theory, and
reading the professional
literature.
This information leads to the
identification of a research
problem, a broad statement
that begins to focus the
direction of study.
The problem is then refined to a
research question, which is
specific and defined.

5. SEVERAL
COMPONENTS
WILL SHAPE THE
QUESTION
evaluation of its importance and
feasibility,
specification of the population to be
studied,
development of a research rationale to
support the question, and a
description of the specific variables to be
studied.

6. Throughout this process, the researcher relies on a comprehensive
review of the literature to provide the background necessary for
decision making.
The research question is then translated into a statement that
reflects the expected outcomes of the study, clarifying the research
objectives in the form of hypotheses or a statement of purpose for
the study.

7. AIM
 framework for developing and refining a
feasible research question,
 to define the different types of variables
that form the basis for the question,
 to describe how research objectives guide
a study,
 to discuss how the review of literature
contributes to this process.

8. FINDING GAPS AND CONFLICTS IN THE
LITERATURE
Professional literature provides the basis for
developing a research problem:
 It will clarify holes in professional
knowledge, areas where we do not have
sufficient information for making clinical
decisions.
 Derives ideas from conflicts in the
literature, when studies present
contradictory findings.

9. EXAMPLE
 the effect of patellar taping for reduction of pain associated
with patellofemoral syndromes.
However, studies have variously shown significant effects,
short-lived effects, or no change in patellar alignment.
Whittingham and colleagues addressed this inconsistent
evidence by designing a randomized trial to investigate the
effect of exercise and taping on pain and function in patients
with knee pain.

10.  identify disagreements due to differences or flaws in study design or
measurement methods.
For example, a systematic review of the efficacy and safety of common
interventions for tears of the rotator cuff in adults showed little evidence to
support or refute the superiority of conservative or surgical interventions. This
supports the need for well designed clinical trials that incorporate consistent
methods for defining interventions and validated outcome measures.
 Research questions may also arise out of data from descriptive studies,
which document trends, patterns or characteristics that can subsequently
be examined more thoroughly using alternative research approaches.
For example, several descriptive studies have documented characteristics of
individuals who have suffered spinal cord injuries. These studies have
provided the foundation for testing new devices to improve function.

11. REPLICATION
 to correct for design limitations
 to examine outcomes with different populations or in
different settings.
 A study may be repeated using the same variables and
methods or slight variations of them.
 Replication is an extremely important process in research,
because one study is never sufficient to confirm a theory or
to verify the success or failure of a treatment.
 We are often unable to generalize findings of one study to a
larger population because of the limitations of small sample
size in clinical studies.

12. EXAMPLE
 Miltner and co-workers studied the effects of constraint-induced movement
therapy in patients with chronic stroke. They cited previous research in American
laboratories showing success of this intervention to improve use of the affected
upper extremity. They were able to replicate these results in Germany, where the
health care system and context of therapy is different than in the United States.

13. THE RESEARCH
RATIONALE
 Once the research problem has been defined, a full review of
literature will establish the background for the research
question.
 This foundation will clarify the research rationale that will
support the research question, guide decisions in designing
the study, and most importantly, provide the basis for
interpreting results.
 The rationale presents a logical argument that shows how and
why the question was developed.
 It provides a theoretical framework by explaining the
constructs and mechanisms behind the question.
 It helps us understand why the question makes sense.
 The research rationale includes references to previous
research as well as logical assumptions that can be made
from current theory. Without a strong rationale, the results of a
study will be hard to interpret.

14. VARIABLES
 Now we must specify what we
want to test.

15. VARIABLES
 Variables are the building blocks of
the research question.
 A variable is a property that can
differentiate members of a group or
set.
 It represents a concept, or factor,
that can have more than one value.

16. EXAMPLE
 if we wanted to compare levels of back pain
between men and women, then pain and gender
are the variables of interest.
 Pain can take on a range of values, depending on
how we measure it, and gender can take on two
"values" (male and female).
 If, however, in another study, we compare the
effects of two different treatments for decreasing
back pain in men, then gender is no longer a
variable. It has only one "value" (male) and is,
therefore, a constant. In this latter example, type
of treatment and pain are the variables of
interest.

17. In descriptive and correlational studies, variables represent the phenomena being examined, and
their measurement may take many forms. The investigator looks at these characteristics one at a
time, describes their values and their interrelationships.
In exploratory and experimental studies the investigator exam ines relationships among two or more
variables to predict outcomes or to establish that one variable influences another.
For these types of studies, research variables are generally classified as independent or dependent,
according to how they are used.

18. INDEPENDENT AND DEPENDENT VARIABLES
A predictor variable is an
independent variable. It is a
condition, intervention or
characteristic that will
predict or cause a given
outcome.
The outcome variable is
called the dependent
variable, which is a response
or effect that is presumed to
vary depending on the
independent variable.

19. VARIABLES IN EXPLORATORY STUDIES
 independent and dependent variables are usually measured together, to determine if they have a
predictive relationship.
Example, researchers have studied the relationship between back pain and age, gender, cognitive status,
ambulatory status, analgesic use, osteoporosis and osteoarthritis in a long-term care population.
 The dependent variable (the outcome variable) was the presence of back pain
 The independent variables (predictor variables) were the characteristics of age, gender, cognitive status
and so on.
 These types of studies often involve several independent variables, as the researcher tries to establish
how different factors interrelate to explain the outcome variable.

20. VARIABLES IN EXPERIMENTAL STUDIES
 involve comparison of different conditions to investigate causal relationships, where
the independent variable is controlled and the dependent variable is measured.
 For instance, researchers have compared the effect of a back class versus usual
medical care to determine if the back class was an effective program for reducing pain
in those with acute low back pain.
 Outcomes included changes in a disability score and a pain scale rating.
 In this example the independent variable is the back class (intervention), and the two
dependent variables are the disability and pain scores (response). A change in the
dependent variables is presumed to be caused by the "value" of the independent
variable; that is, the dependent variable is a function of the condition of the
independent variable.

21. Comparative studies can be designed with more than one independent variable.
We could look at the patients' gender in addition to intervention,
for instance, to deter mine if effectiveness of a back class is different for males and females.
We would then have two independent variables: type of intervention and gender.
A study can also have more than one dependent variable. In the previously mentioned study,
researchers measured both disability rating and pain.

22. LEVELS OF THE INDEPENDENT VARIABLE
In comparative studies, independent variables are given "values" called levels.
The levels represent groups or conditions that will be compared. Every independent variable will have
at least two levels. Dependent variables are not described as having levels.
For example, in the study comparing a back class and usual care, the independent variable of
"intervention" has two levels: back class and usual care. If the study had included additional
interventions, such as physical therapy or bed rest, it would have changed the number of levels of the
intervention variable, not the number of variables.

23. OPERATIONAL
DEFINITIONS
defines a variable according to its
unique meaning within a study
The operational definition should
be sufficiently detailed that
another researcher could replicate
the procedure or condition.

24.  Independent variables are operationalized according to
how they are manipulated by the investigator
 For example, in the study comparing a back class and
usual care, an operational definition for the
independent variable "back class" should include the
number of sessions, the type of training and materials,
expectations of compliance, who will teach the class
and so on. The subjects' activities and other treatment
specifications should be included. We also need to
describe the control group's activities of usual care.
 Operational definitions for independent variables must
differentiate the various levels of the variable.

25. Dependent variables are operationally defined by describing the method of
measurement, including delineation of tools and procedures used to obtain
measurements.
A variable like "low back pain" could be defined operationally as the score on a
visual analog scale (VAS), reflecting the magnitude of pain at a particular time of
day under specific activity conditions.
An individual reading this definition should be able to know precisely how the
variable "pain" could be interpreted in this study.