Transcript Research Design - University of West Alabama

Research Design
Research is based on Scientific Method
Propose a hypothesis that is testable
Objective observations are collected
Results are analyzed in an unbiased manner
Conclusions proposed are based on the results
of the study and previous knowledge
Identifying the problem
Perform a Literature Review
Formulate the Purpose
Design Procedures
Review of Study
When using human subjects, protecting them is of up
most importance. In the research world, this is
accomplished by independent review of the
research proposal and methods by an IRB
(Institutional Review Board) . Those conducting
research outside of institutions should refer to
administration or independent reviews for this
process and use the Nuremburg Code, a version of
the Declaration of Helsinki to govern choices.
Once approved, if design changes or methods are
altered, the design must be resubmitted.
Collect Data
What Are Statistics?

The means by which quantitative data are
organized, analyzed and interpreted. Only
through statistical application may scientific
data be treated and made meaningful
Interpret Findings and Form
Conclusions
Publish Your Answer
Evaluating a Study
There are many reasons many
professionals choose not to perform
research. But being able to evaluate
publications is of paramount
importance.
Evaluating a study
Literature Review
 Design
 Assignment
 Assessment
 Analysis
 Interpretation
 Extrapolation

Lit Review
Separated subjective opinions and untested
theories form research finding
 Was investigator objective in listing results
form studies that refuted as well as stated
his/her hypothesis
 Were previous studies reported correctly
 References are current
 Varity of journal, sources, and related
professions consulted

Proper Design..
Purpose of study defined
 Hypothesis correctly stated
 Was study type appropriate to the question
being answered

Does the assignment of experimental and
control groups demonstrate integrity?
Researches try to make both groups
identical except for characteristics under
question
 Selection Bias occurs when an investigator
unintentionally introduces factors into the
selection of the study that can predetermine
the outcome.

Assessing Results of a Study

Research must define the outcome that is to
be measured and that outcome must meet
these criteria
– Use a measure of outcome that is appropriate
the the question to be answered
– Measurements of the outcome must be precise
– Measurement of the outcome must be complete
– The outcome of the study must not be
influenced by the process of observation
Analysis




Researcher compares result of control and
experimental groups.
Researcher identifies and adjusted for
(confounding variables) factors other than those
being studied
After adjustments, statistical tests are performed
Research reports the size of the differences and the
degree of overlap in differences presented
Interpretation

Deciding the clinical usefulness of the
results obtained or determining whether a
cause and effect relationship has been
established
Extrapolation


(Inference) involves obtaining the meaningful
study for the larger, outside population.
Four types of extrapolation errors when
extrapolating to larger population
– Beyond data range (longer duration of experiment will
produce same effects in subjects)
– Inferring conclusions about individuals from population
data
– Error resulting from unappreciated factor in new
population
– Errors in going from study population to general
population
19 Questions to ask
1.
2.
3.
4.
5.
6.
Was the purpose of the study sufficiently
designed?
Were the study hypothesis clearly stated?
Was an appropriate question being asked
What was the study population? Was it adequate
composition and size to answer the study
questions?
Was the assignment of patients to study and
control groups proper? Could selection bias have
occurred?
Were the study and control groups comparable
with respect to the characteristic other that the
study factors(s).
7.
8.
9.
10.
11.
12.
Was the measure of outcome appropriate to the
study aims?
Was the measure of outcomes precise?
Was the measure of outcomes complete?
Did the process of observation affect the
outcome?
Were the results adjusted to take into account the
effect of the possible confounding variables?
Was a significant test properly performed to
assess the probability that the difference was due
to chance?
13.
14.
15.
16.
17.
18.
19.
Was the proper measure of the size of the
difference presented?
Was a proper Measure of the degree of overlap
of the difference presented?
Did the investigators properly reject or fail the
null hypothesis?
Interpreting the meaning of any relationship,
was the clinical concept of cause and effect
properly applied?
Did the investigators stay within the limits of the
data hen extrapolating the results?
If the investigators extrapolated from population
data to individual data, did they committee an
ecological fallacy?
Did researchers take into consideration