NURS301 Lecture Notes - Lecture 2: Analysis Of Variance, Evidence-Based Medicine, Central Tendency
Quantitative
Qualitative
Experimental
(RCT)
Quasi-
experimental
Non-experimental
Mixed methods
Systematic Review
Meta-analysis
Research
purpose
• Test
hypotheses that
describe
phenomena
• Test
relationships
between
variables
• Answer
questions or
solve problems
• Explain cause-
and-effect
interactions
• Hypothesis
present
• Research
questions
about human
experiences of
health
• Help us
understand the
complexity of
human health
experience,
create
instruments
and develop
theory
• No hypothesis
• Random
assgn. to tx or
control group
• Control
• Manipulation
of variable
• Removes
sources of
bias; gold
standard
• Construct a
picture of a
phenomenon
at one point
or another
• Explore
people,
places,
events,
situations as
they occur
naturally
• Test
relationship /
differences
among
variables
• Mixture of
qualitative and
quantitative
methods to
answer
research
question
• Used in policy
development,
organization
studies and
program
evaluation
• A form of
triangulation
• Special kind of
literature
review
• To find,
appraise and
synthesize all
empirical
evidence that
fits pre-
determined
eligibility
criteria using a
systematic
approach that
aims to
minimize bias
• Rigorous
methods to
identify,
critically
appraise, and
synthesize
primary studies
• Best evidence
• Critical
thinking before
implementing
into clinical
practice
• Highest level of
evidence
• Strict scientific
process that
synthesizes the
findings from
several separate
studies in a
specific area and
statistically
summarizes the
findings to
obtain a precise
measure of the
effect
• Statistically
combine
findings from
multiple studies
focused on
similar variables
find more resources at oneclass.com
find more resources at oneclass.com
Design
• Apply different
levels of
control
• Includes:
Participant
Observations
Measurement
of time
Selection of
subjects
Role of
investigator
• Consider:
Objectivity
Accuracy
Feasibility
Control
Homo sample
Constancy
Manipulation
Randomization
(baseline test,
don’t work if
sample<100)
• Internal validity
(If independent
variable or
some other
factor affected
change in
dependent
variable)
1. History
2. Selection
3. Maturation
4. Testing
5. Mortality
6.
Instrumentation
by random
selection
• Grounded
theory
Inductive
theory from
lived
experiences
about basic
social
processes
• Case studies
Investigates
contemporary
phenomenon
• Historical
Systematic
compilation of
data to
describe past
event/people,
root in
philosophy,
art, science
• Ethnography
Describe
cultural
groups, root in
cultural
anthropology
• Phenomenolo
gy
Understand
phenomenon
from people
experiencing
it, root from
philosophy
• Participatory
action
Find solutions
as a
community
• Cons:
Natural setting
• True
experimental
design
• Solomon four-
group design
• After-only
design
• Pros:
Good for
testing
cause/effect
Highest level
of evidence
for single
studies
• Cons:
Not all
eseach ’s
are amenable
to
experimental
manipulation /
randomization
Subject
mortality
Diff. logistics
in field
settings
Hawthorne
effect
• Pros:
Practical,
feasible
Some
generalizability
• Cons:
Not good for
cause/effect
May not be
able to
randomize
• Survey
studies
Descriptive
Exploratory
Comparative
• Relationship
or difference
studies
- Correlati
onal
- Develop
mental
Cross
sectional
Longitudi
nal/prosp
ective
Restrosp
ective/ex
post
facto
• Pros:
Helps develop
knowledge
base
Useful in
forecasting /
making
predictions
Imp. when
randomizatio
n,
manipulation
and control
not possible
• Timing?
Concurrent or
sequential
• Weight of
each method
Equal vs
unequal
• How to mix?
Merge during
interpretation/a
nalysis, embed
in design as
secondary
piece, connect
data so one
leads to /
builds on other
• Triangulation
Obtain
different info
on a single
topic
“+” concurrent
“A” primary
Difficult if
findings don’t
converge
• Embedded
One set of data
is secondary to
other
“()” embedded
• Explanatory
Two phase /
sequential
where one set
of qualitative
data builds on
initial
quantitative
results
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Construct a picture of a phenomenon at one point or another, explore people, places, events, situations as they occur naturally, test relationship / differences among variables. Experimental (rct: random assgn. to tx or control group, control, manipulation of variable, removes sources of bias; gold standard. Design: apply different, grounded, true, pros: levels of control. Internal validity (if independent variable or some other factor affected change in dependent variable: history, selection, maturation, testing, mortality. May not be able to randomize experimental design: solomon four- group design, after-only design, pros: Highest level of evidence for single studies: cons: Not all (cid:396)esea(cid:396)ch (cid:395)"s are amenable to experimental manipulation / randomization. Inductive theory from lived experiences about basic social processes: case studies. Systematic compilation of data to describe past event/people, root in philosophy, art, science: ethnography. Describe cultural groups, root in cultural anthropology: phenomenolo gy. Understand phenomenon from people experiencing it, root from philosophy: participatory action.