Lecture 10 - Designing Experiments, Blocking/Replication/Randomization, Causation, Sampling Flashcards

1
Q

Which of the following best distinguishes an experimental study from an observational study?

A) Unlike observational studies, experimental studies must be run by statisticians.
B) In an experimental study, the investigator assigns some treatment to units involved in the study to observe the response.
C) An experimental study requires more people in the study compared to an observational study.
D) All of the above

A

B) In an experimental study, the investigator assigns some treatment to units involved in the study to observe the response.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Which of the following is an observational study?

A) Researchers study the smoking habits of a group of patients in a hospital and record the incidence of lung cancer.
B) Doctors measure the temperature of patients in the ICU to see if the risk of sepsis increases with the increase in temperature.
C) Researchers record the eating habits of children over time to study the association between their diet and height.
D) All of the above

A

D) All of the above

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Principles of Experimental Design

Example: Sleep and Grades

A

We know that age and IQ are controlled. This means that changes in quiz grade with respect to sleep cannot be due to age or IQ.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Blocking

A
  • We want the only difference between the pair to be the treatment. Thus, if
    the response changes, we can say it’s due to the treatment
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Replication

A
  • There could be extra variables we did not control for when we blocked.
  • By replicating and randomizing, we can spread out confounding or lurking variables between the different treatment groups.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Randomization

A
  • There could be extra variables we did not control for when we blocked.
  • By replicating and randomizing, we can spread out confounding or lurking
    variables between the different treatment groups.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Un-Randomized Situation

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Randomized Situation

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Purpose?

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Activity: Read the following scenario and answer the questions

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Question

In this study, the unimpaired vision group is a:

A) Confounder
B) Control
C) Block
D) Match
E) None of the above

A

B) Control

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Question

In this study, the amount of prior driving experience is a:

A) Confounder
B) Control
C) Block
D) Match
E) None of the above

A

A) Confounder

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Causation AND
Observational Studies

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Activity: Design the following experiment.

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Causation AND
Experimental Studies

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Samples

Population vs Sample

A
17
Q

Statistical Inference

The procedure by which we reach a conclusion about a population based on the information contained in a sample that has been drawn from that population

A
  • To produce precise inference, a sample needs to be as representative
    as possible of the population of interest
  • If the sample is not a good representation of the population, sample error may occur making the statistical inferences invalid
18
Q

Sampling Designs/Methods

Non-Probability Sampling

A
19
Q

Sampling Designs/Methods

Probability Sampling

A
20
Q

Sampling Designs/Methods

Simple Random Sampling (SRS)

A

size 𝑛 consists of 𝑛 individuals from the population chosen in such a way that every set of 𝑛 individuals has an equal chance to be the sample actually selected

21
Q

Example

Simple Random Sampling (SRS)

A
  • Picture a large amount of poker chips in a bowl with information written on each one.
  • Reach in and randomly select 𝑛 chips.
  • This represents our SRS.
  • In practice, we sample without replacement (once a chip is selected, it’s not returned to the bowl).
22
Q
A