Professional Documents
Culture Documents
RELATIONSHIP BETWEEN
VARIABLES
Ketut Suega
Unit Litbang dan ICE-EBM
FK Unud/RSUP Sanglah
Denpasar Bali
Definitions
• Variable: “any entity that can take on a variety of different
values” (Wrench et al, 2008, p. 104)
– gender
– self-esteem
– managerial style
– stuttering severity
• attributes, values, and levels are the variations in a variable
– Attribute: political party:
– Value: Democrat, Republican, Independent, etc.
– Attribute: Self-esteem
– Level: High, Medium, Low
What is a Variable?
5 10 15 20 25
Height of a ramp (cm)
interchangeability of independent and
dependent variables
• The same concept or construct could serve as the
independent variable in one investigation, and the
dependent in another.
• example: “fetal alcohol syndrome” (FAS)
– As an independent variable: RQ: Does severity of
FAS correlate positively with language delay in
infants?
– As a dependent variable: RQ: Does the amount of
maternal alcohol use correlate positively with the
severity of FAS in infancy?
TYPE OF VARIABLES
TYPE OF VARIABLES
TYPE OF VARIABLES
TYPE OF VARIABLES
TYPE OF VARIABLES
TYPE OF VARIABLES
Scale of Variables
• Discreetly (catagorical) • Continously (numeric)
– Nominal – Interval
– Ordinal – Ratio
Varieties and types of variables
• Discrete variables
– Nominal variables: distinct, mutually – Dichotomous variables:
exclusive categories • true/false, female/male,
• religions; Christians, Muslims, democrat/republican
Jews, etc. – Ordered variables: mutually
• occupations; truck driver, exclusive categories, but with an
teacher, engineer order, sequence, or hierarchy
• marital status; single, married, • fall, winter, summer, spring
divorced • K-6, junior high, high school,
– Concrete versus abstract variables college
• concrete; relatively fixed,
unchanging
– biological sex
– ethnicity
• abstract; dynamic, transitory
– mood, emotion
– occupation
Varieties and types of variables--
continued
• Continuous variables:
• include constant increments or gradations, which can
be arithmetically compared and contrasted
– IQ scores
– self-esteem scores
– age
– heart rate, blood pressure
– number of gestures
Relationships among variables
• Differences
– Differences in kind, degree
• Relationships (correlations)
– Positive correlation
– Negative correlation
– No or neutral correlation
Relationships among variables
chances
bias
confounding
causal
If exposure X is If exposure X is
associated with outcome Applying guidelines associated with outcome
Y…..then how do we decide for Y…..then how do we
if X is a cause of Y causal inference decide if X is a cause of
Y
Is this association causal?
• Two-stage process:
• Stage I:
– Consider alternative “non-causal explanations” for the association
Selection or measurement
bias
No
Stage I
Confounding
No
Apply Guidelines
Chance for Causal Inference
Probably Not
Could it be causal?
Stage II
BIAS
Dependent Variable
Independent Variabel = dependen
= independent Mediating = efect
= predictor Variable = outcome
= risk = event
= causa = response
External Confounding
External
Variabel Variabeles Variabel
D
A
External
Variabel External
Variabel
B C
Independent Variabel Dependent
Variabel antara Variable
A Confounding
B Independent Dependent
Variabel M Variable
Not Confounding
C Independent Dependent V
Variabel Variable
?
Coffee PJK
Smoking
Candy Dental carries
Snacks
Oral hygiene
CONTROLING CONFOUNDING
FACTORS
BY IDENTIFICATION
BY ELIMINATING
IDENTIFICATION
REVIEWING ARTICLES
LOGICAL
EXPERIENCE
ELIMINATION
• ANALYTIC PROPERTIES
– STRATIFICATION
– MULTIVARIATE
• STUDY DESIGN
– RESTRICTION
– MATCHING
– RANDOMIZATION
Dr. Yoga Nathan
Senior Lecturer in Public Health
GEMS UL
“In what circumstances can we pass from an observed association to a
verdict of causation? Upon what basis should we proceed to do so?”
Strength Plausibility
Consistency
Coherence
Specificity
Experimental
evidence
Temporality
Dose-response Analogy
Consistency
• Repeated observation of an association in studies conducted on
different populations under different circumstances
• If studies conducted by….
– different researchers
– at different times
– in different settings
– on different populations
– using different study designs
……all produce consistent results,
this strengthens the argument for causation
Epidemiological studies (1 - 14)
study
Consistency
• Repeated observation of an association in studies conducted on
different populations under different circumstances
• If studies conducted by….
– different researchers
– at different times
– in different settings
– on different populations
– using different study designs
……all produce consistent results, this strengthens the argument for
causation
Outcome
Exposure TIME Normal Cancer
lung
Temporality – British Doctors Cohort Study
Lung Ca.
• This refers to the necessity for the exposure to
precede the outcome (effect)Smokers
in time no
Lung Ca.
• Any claim of causation must involve the cause
Lung Ca.
preceding in timeBritish
40,634 the presumedExeffect
Population Doctors Smokers
no
• Easier to establish in certain study designsLung Ca.
– Prospective cohort study
Lung Ca.
Non
Lack of temporality rules out causality
Smokers no
Lung Ca.
Time
Exposure TIME Outcome
Temporality
• This refers to the necessity for the exposure to
precede the outcome (effect) in time
• Any claim of causation must involve the cause
preceding in time the presumed effect
• Easier to establish in certain study designs
– Prospective cohort study