Lec 5-8. Measures of association

Question 1

What is the counterfactual theory?

Answer 1

All else being equal (in the same group) the outcome if the exposure didnt occur. In other words the best outcome of comparing groups would be comparing the group while they got exposed to the same exact group like they they never exposed. Its impossible.

Question 2

What is exchangeability?

Answer 2

Comparability with respect to all other determinants of outcome. How the similar the two groups are.

Question 3

Define absolute differences

Answer 3

Subtracting frequecies (counts)

Question 4

Define Relative differences

Answer 4

Division (ratio) of frequencies (counts) or proportions

Question 5

What is the difference between relative and absolute differences?

Answer 5

Absolute differences tend to be smaller and absolute differences deal with subtraction

Question 6

Define risk

Answer 6

Proportion, simple part over whole. Outcome whether a person was exposed or not exposed.

Question 7

What is the Incidence Risk (IR) equation for risk in exposed?

Answer 7

A/(A+B)

Question 8

What is the Incidence Risk (IR) equation for risk in non-exposed?

Answer 8

C/(C+D)

Question 9

What is NNT/NNH?

Answer 9

of pts needed to be treated to receive the state benefit/harm.

Question 10

What is the equation for NNT/NNH?

Answer 10

1/Absolute risk reduction (ARR)

Question 11

If the NNT is 3.03. How many pts need to be treated to recieve that benefit?

Answer 11

4

Question 12

What value would we like our NNT to be?

Answer 12

Small

Question 13

What value would we like our NNH to be?

Answer 13

large

Question 14

What is the abbreviation for Risk Ratio?

Answer 14

RR

Question 15

What is the equation for Risk Ratio?

Answer 15

(Risk outcome in exposed/Risk outcome in unexposed)

Question 16

If the Ratio is >1.0 then the event/outcome is

Answer 16

more likely to occur. Increased risk.

Question 17

If the Ratio is <1.0 then the event/outcome is

Answer 17

Less likely to occur. Decreased probability

Question 18

If the Ratio is 1.0 then the event/outcome is

Answer 18

EQUALLY likely for both groups

Question 19

If the Ratio is >2.0 then the event/outcome is

Answer 19

“x” times greater. Ex: OR=6.19, thus the comparator group is 6.19 times greater

Question 20

When interpreting Ratio’s (RR/OR/HR) what three things do we look for?

Answer 20

Group comparison orientation, direction of words, and magnitude.

Question 21

What is an example of group comparison orientation?

Answer 21

Ramipril vs. placebo OR placebo vs. Ramipril

Question 22

What is an example of Direction of words?

Answer 22

increased/decreased

Question 23

What is an example of Magnitude?

Answer 23

80% (1.8 times) or 20% (0.8 times)

Question 24

When looking at the CI for Ratios (RR/OR/HR) if both values of the interval are on the same side of 1.0, what does that mean?

Answer 24

It is always Significant

Question 25

What is the definition for odds?

Answer 25

Frequency of an outcome occurring vs NOT occurring.

Question 26

What are the equations for Odds?

Answer 26

in cases (A/C), in controls (B/D)

Question 27

What is the equation for Odds ratio?

Answer 27

[(AxD)/(BxC)]