Lecture 6

Question 1

For a binary outcome, what are 2 ways of Quantifying the strength of association?

Answer 1

1. Difference
        • risk diff
               - "their risk has ↑ by X"
        • rate diff
2. Ratio
        • risk ratio (relative risk)
                -"their risk ↑ X fold"
        • rate ratio
        • odds ratio

Question 2

what is an indicator of precision?

Answer 2

Confidence interval

• do ration btwn low vs high end of interval
• • higher the number the less precise

Question 3

what is an indicator of statistical significance?

Answer 3

p-value

Question 4

What is the range used for difference?

What implies no association?

Answer 4

Neg infinity to Positive infinity

0 = no assoc
• farther away from zero, stronger assoc

Question 5

What is the range used for Ratio?

What implies no association?

Answer 5

ZERO to indinity

1 = no assoc
> 1 = + assoc
< 1 = - assoc

Question 6

Odd ratio

Odds

Answer 6

Odds ratio
• relative measure based on “odds”

Odds
• ratio of probability of an event occurring : the probability of it not occurring
• Range 0-infinity

Question 7

What do you want to asses if you are trying to see if there is a causal relationship?

Answer 7

ABCs
• Is the observed ASSOCIATION due to:
   • Bias
   • Confounding
   • Chance
   • Cause

Question 8

How do you determine if there is an association?

Answer 8

• randomized controlled trial

* difference OR ratio of effect size btwn Tx groups

Question 9

Bias

Answer 9

Aka Systematic Error
• Selection
• Information
• Specification

• • Not removable by statistics!
• • STUDY DESIGN – reduces bias

Question 10

Selection bias

Answer 10

• loss to follow-up

- control selection is dependent of exposure of interest

Question 11

Information bias

Answer 11

• misclassification
• recall bias
• not blinded to as Tx

Question 12

Specification bias

Answer 12

• wrong statistical models

Question 13

Confounding

Answer 13

• Lack of comparability btwn study groups
• results in bias in estimation of the effect of interest
• • adjustable statistically
• • ↓ w/ study design

Question 14

Which can be adjusted w/ statistical methods… Confounding or Bias?

Answer 14

Confounding

Question 15

What must a confounding variable be?

Answer 15

1. A cause of the outcome
   - (independent of exposure)
2. Correlated w/ exposure in study population
3. Not affected by the exposure or outcome

Question 16

Chance

Answer 16

aka Random Error
• statistics quantify probability that association is due to chance
• assumes random sampling & randomization

Measures
• P-value – small sample –> ↑ ↑ p-value
• confidence interval – Wide = non-precise

Question 17

Valid result

Answer 17

= unbiased result

• measuring exactly what we WANT to be measuring

Question 18

Precise result

Answer 18

= small variability

Question 19

Factors affecting validity

Answer 19

• bias

* confounding

Question 20

Factors affecting precision

Answer 20

• Random variation/error
• sample size
• Study design (efficiency)

Question 21

Hill’s guideline

Answer 21

Used to determine causal vs non-causal associations

1. Strength of association
2. Consistency
3. Specificity
4. Temporality <—-MUST be there.
5. Biological gradient (dose-response)
6. Plausibility
7. Analogy

Question 22

Can a weak association be causal?

Answer 22

Yes, but a strong association is more likely causal

Question 23

Does a strong association indicate cause?

Answer 23

NO! (think literbox example)

Question 24

Consistency

Answer 24

Repeated observations of an association in diff populations under diff circumstances
• meta-analysis = good for assessing consistency

Question 25

Does lack of consistency R/O causal association?

Answer 25

• No, some effects are produced under unusual circumstances

Question 26

Specificity

Answer 26

One exposure –> one effect (not multiple)

• not often the case

Question 27

Temporality

Answer 27

ONLY necessary criteria

• cause precedes effect

Question 28

Biological gradient

Answer 28

Presence of monotonic dose-response curve (unidirectional)

Question 29

Analogy

Answer 29

Finding similar results elsewhere

Question 30

Multifactoral etiology

Answer 30

various factors interact to result in disease

• think of dz as being an ECOLOGICAL problem