Académique Documents
Professionnel Documents
Culture Documents
A
scienHc
invesHgaHon
An
orderly
system
A
tool
for
growth
A
problem-solving
acHvity
The
problem
is
coming
up
with
a
problem!
7/8/10
Example
A
What
do
I
want
to
nd
out
and
how?
Who
is
it
for?
What
do
I
expect
to
happen?
(Punch,
2000
and
Kumar,
1996)
IntroducHon
(Creswell, 1994)
Example B
Research
quesHons
general
specic
Signicance
LimitaHons
and
delimitaHons
(if
appropriate)
Consent,
access
and
parHcipants
protecHon
References
Appendices
(Punch,
2000,
p.
67)
Example C
7/8/10
Undergraduate
Clinical
Epidemiology
Course
Improves
Clinical
Prac,ce
7/8/10
2
Main
Purposes:
1. To
akract
readers
2. To
tell
them
what
to
expect
Must
demonstrate
to
readers/reviewers
that
they
are
reading
the
work
of
a
professional
Should
be
short
1-2
pages
Funnel structure
7/8/10
Importance:
Importance:
Shows
that
research
idea
has
a
sound
scienHc
or
biologic
raHonale
7/8/10
7/8/10
1. Bibliography
2. Informed
Consent
Form
3. Case
Report
Form
or
Data
CollecHon
Form
4. Flowchart
of
AcHviHes
5. QuesHonnaire
6. Dummy
Tables
Qualitative
Quantitative
Analytic
Descriptive
Observational
Crosssectional
Cohort
Retrospective
Casecontrol
Prospective
Interventional
Natural
experiment
RCT
Quasiexperiment
7/8/10
Cross-sectal
Case- control
Cohort
Classification
of population
Population free
from disease
Rare condition
N/A
Best
N/A
To determine a
precise risk
Gives relative
prevalence
Only estimate
possible
BEST
Sample
represented
Survivors at
point in time
Uncertain
Non-diseased
N/A
BEST
Temporal
sequence
Retrospective,
relship revealed
at point in time
Retrospective
Prospective or
retrospective
N/A
N/A
Function
Describes assn
bet. exp & dis
simultaneously
Compares
prevalence of exp.
in cases & control
Compares
incidence in
exposed & unexp.
Attrition
usually
minimal
N/A
Evidence of
causality
Only suggestive
Needs more
careful analysis
STRONG
LEAST
EXPENSIVE
MOST
EXPENSIVE
Bias
May be difficult
to manage
EASY TO
MANAGE
Cross-sectal
Case- control
Cohort
Cross-sectal
Case- control
Cohort
Time relationship
ED
ED
ED
Data collection
Slice of time
Retrospective
Forward
Subject
recruitment
Exposed &
unexposed,
look at old &
new cases at
the same time
Exposed &
unexposed &
prospective
surveillance of
new cases
Uses
Reliability &
validity,
epidemics,
prevalence
studies
Rare exposures,
short duration
diseases,
incidence
studies
Levels of bias
++
+++
++
+++
Strength of results +
BEST
In between