Study Designs

Question 1

Correlation

Answer 1

when 2 things appear related

Question 2

Causation

Answer 2

one thing (cause) is causing another thing (effect)

Question 3

Confounding Variables

Answer 3

factors that can distort relationship between variables, may be inaccurate or misleading

Question 4

Causal relationships can be

Answer 4

direct or indirect

Question 5

direct relationships

Answer 5

one variable directly causes another without an intermediate
e.g. smoking directly causes lung cancer

Question 6

indirect relationship

Answer 6

factor causes disease but only through an intermediate of steps

Question 7

why should we determine a causal relationship

Answer 7

to predict, influence and understand the outcome

Question 8

Adjustment

Answer 8

used to allow populations to be compared when they are not comparable

Question 9

Age adjustment

Answer 9

compare different populations or subgroups

Question 10

age standardization

Answer 10

summary measure that reflects what the rates would be at the same age distribution

Question 11

comparing mortality

Answer 11

populations differ in characteristics, should adjust for factors that greatly affect mortality.

e.g. age is the most important predictor of mortality

Question 12

explain significance of sample adjustment

Answer 12

bad samples can lead to misleading comparisons, inaccurate risk assessment, limited comparability

Question 13

Historical randomized trial

Answer 13

performed by James Lind, observed vitamin C deficiency in sailors (scurvy)

Question 14

Process of randomized clinical trial

Answer 14

1. Participants
2. Randomly assign participants to treatment or placebo (control)
3. compare results

Question 15

Purpose of randomized clinical trial

Answer 15

evaluating new treatments

Question 16

Selection of subjects

Answer 16

study must be replicable, criteria for determining who will or will not be included must be detailed. no room for bias

Question 17

subject blinding

Answer 17

when study subjects do not know which control they are assigned

Question 18

Observer blinding

Answer 18

observers or data collectors do not know where control was assigned to prevent bias

Question 19

Placebo

Answer 19

treatment or response that appears real but does not do anything. may elicit positive result

Question 20

Crossover trials

Answer 20

type of clinical trial where subjects are randomized to new treatment, after period of time switched