Lecture 8: Casuality, bias and confounding Flashcards

(31 cards)

1
Q

What is a casual inference?

A

determining the effect of a phenomenon part of a wider system

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

What is the name given to what would have happened in a study if we’d done things differently?

A

counterfactual

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

What is another name for an estimate of a counterfactual?

A

Control

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

What term is used to describe groups of people in a study who are comparable on average?

A

exchangeable

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

What is the easiest way to ensure exchangeable study groups?

A

randomisation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What is random sampling error?

A

-The random error in our population estimate(s) that results from chance fluctuations in the profile of our sample

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Is error caused by random or non-random factors?

A

non-random

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

What is bias?

A

a non-random, systematic error

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

What is precision?

A

a measure of how close a series of measurements are to one another

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

What is accuracy?

A

how close a measurement is to the true value

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

What is confounding bias?

A

Distortion of the causal association between two variables, due to a common shared cause (a confounder)

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

What is a confounder?

A

a third variable that influences an association without being measured

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

What is conditioning?

A

the process of reducing confounders by examining like-for-like participants (grouping participants into exposed and non-exposed groups)

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

Give 3 forms of conditioning confounders:

A

1) restriction

2) stratification

3) covariate adjustment

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

Describe restriction as a form of conditioning:

A

restricting the sample to a single value of the confounder

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

Describe stratification as a form of conditioning:

A

calculating category-specific effects for different levels of the confounder

17
Q

Describe covariate adjustment as a form of conditioning:

A

adjusting the cofounders in a regression of the association

18
Q

What causes selection bias?

A

a systematic difference between those selected into a study sample and those that were not selected

19
Q

Give 3 types of selection bias:

A

1) sampling bias

2) participation bias

3) attrition bias

20
Q

What is sampling bias?

A

a failure to sample evenly across the population (not generalisable)

21
Q

What is participation bias?

A

people having different preferences or opportunities to participate in research

22
Q

What is attrition bias?

A

a loss of participants from the study which may be unbalanced by the exposure

23
Q

What causes information bias?

A

a systematic error in reporting, measurement or the recording of error

24
Q

Give 3 types of information bias:

A

1) response bias

2) recall bias

3) measurement bias

25
What is response bias?
people responding in inaccurate or untruthful ways by wanting to give the right answer or downplaying undesirable traits
26
What is recall bias?
people having different abilities to remember past information
27
What is measurement bias?
a measurement that under or over reports systematically
28
What causes experimenter bias?
the behaviours and actions of the experimenter whether conscious or unconscious
29
Give two types of experimenter bias:
1) confirmation bias 2) results bias
30
What is confirmation bias?
where findings we expect are more likely to be accepted and those that aren't are refuted
31
What is results bias?
being more likely to chase positive associations, seek novel results or publish positive results