Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

Epidemiology > Validity of studies. > Flashcards

Validity of studies. Flashcards

1
Q

What is the overall goal of epidemiology?

A

To identify a valid and precise estimate between an exposure and outcome.

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

Once interpreted a statistical association may be found not to be causal. What may it be instead?

A

A systematic error (bias), confounding issue or random error.

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

What is the better way to refer to ‘validity’ in regards to epidmiological studies?

A

Internal validity.

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

What needs to be absent for a study to have internal validity?

A

Systematic error (bias).

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

If a study is precise what is absent?

A

Random error.

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

What does the study result need to be in order for the result to be defined as precise?

A

Repeatable.

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

A study that has high validity also has high precision. True or false?

A

False.

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

What can external validity also be known as?

A

Generalisability.

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

For a result to be valid from an epidemiological study what must be absent?

A

Bias.

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

What three things effect internal validity?

A
  1. Selection bias.
  2. Information bias.
  3. Confounding.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What is this the definition of ‘ Any trend in the collection, analysis, interpretation, publication or review of data that are systematically different from the truth’.

A

Systematic error (bias).

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

Bias is more likely to come from the study design stage than any other stage of the study. True or false?

A

False. It is also likely to arise when the study is being conducted (if it is being conducted inappropriately).

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

Bias is caused by systematic error. What is chance caused by?

A

Random error.

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

What sort of errors can cancel themselves out long term as the sample size increases?

A

Errors from chance.

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

In very large samples errors from bias can cancel. True or false?

A

False.

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

What sort of results can chance cause?

A

Imprecise results.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

What causes selection bias in cohort studies?

A

Differential loss to follow up.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Is selection bias more likely to occur in case control or cohort studies?

A

Case control.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

What does information bias cause?

A

Misclassification.

20
Q

What two types of biases can lead to information bias?

A
  1. Interviewer bias.

2. Recall bias.

21
Q

What causes selection bias (by definition)?

A

When the chance of an individual getting into the study sample is related to both the exposure and study outcome.

22
Q

What is case control status another term for?

A

Outcome.

23
Q

What type of bias occurs in hospital based case control studies? Why does this occur?

A

Berkson’s bias.

Individuals are more likely to be exposed to the exposure under consideration compared to the general population.

24
Q

What does Berkon’s bias do to the odds ratio?

A

It underestimates it.

25
Q

What is the Healthy Worker Effect?

A

An effect that occurs in occupational cohort studies when a group of individuals exposed to an occupational hazard are compared to an unexposed sample in a general population.

26
Q

Why does differential loss of follow up lead to selection bias?

A

Individuals who are lost to follow up tend to have different probabilities of the outcome compared to those who remain throughout the study.

27
Q

If sampling fractions/ response rates differ by case-control status and not exposure status does bias occur?

A

No.

28
Q

If sampling fractions/ response rates differ by exposure status and not case-control status does bias occur?

A

No.

29
Q

When does selection bias occur in a study in terms of sampling fractions/response rate?

A

When sampling status and response rate differs depending on exposure status and case/control group.

30
Q

Does selection bias occur towards or away from the null? (or both)?

A

Both.

31
Q

If selection bias is not prevented at the study design stage it is unlikely to be fixed at the data analysis stage?

A

True.

32
Q

What three things can be done to minimize selection bias?

A
  1. Select cases and controls from the same population.
  2. Controls have the same prevalence of exposure to the overall population.
  3. Non response kept to a minimum.
33
Q

If exposure status is misclassified what measurement becomes biased?

A

OR.

34
Q

What causes information bias?

A

The case/ control status of an individual not being masking resulting in the interviewer obtaining differential amounts of exposure information for cases and controls.

35
Q

Quality control activities can prevent information bias. What else can? (3 things).

A
  1. Validity studies.
  2. Masking case-control status.
  3. Masking study hypothesis.
36
Q

Quality control activities can prevent information bias. Name three examples of these quality control activities?

A
  1. Manual of operation.
  2. Monitoring of information collecting procedures.
  3. Standardization of data collection procedures.
37
Q

What three things can prevent recall bias?

A
  1. Verification of exposure (eg medical records).
  2. Objective biomarkers.
  3. Diseased control group.
38
Q

What is the definition of recall bias?

A

Differentially inaccurate recall of past exposure by cases and controls status.

39
Q

What is non differential misclassification?

A

Errors in the determination of an individuals exposure status that occur the same amount in cases and controls ( or reverse).

40
Q

Does non differential misclassification bias result in bias towards or away from the null?

A

Towards.

41
Q

What is differential misclassification bias?

A

Errors in the determination of an individuals exposure status that occur in different amounts in cases and controls ( or reverse).

42
Q

Does differential misclassification bias result in bias towards or away from the null?

A

Either.

43
Q

What does this describe?

‘Errors in exposure classification depend on case-control status’.’

‘Errors in outcome classification depend on exposure status’.

A

Differential Misclassification.

44
Q

What four things can prevent misclassification bias?

A
  1. Avoid recall or information bias.
  2. Use validated reliable assays for biomarkers and assay cases and controls together.
  3. Use gold standard diagnostics for outcome assessment.
  4. Screen controls for subclinical disease.
45
Q

What causes detection bias?

A

A medically related exposure leads to the detection of the study outcome, particularly in diseases with a high proportion of subclinical cases.

46
Q

What does detection bias do to the odds ratio?

A

Exaggerate it.

47
Q

Once you have checked that your study results have not occurred due to bias what must you then do?

A

Check the risk ratio, P-values 95% CI’s and power of the study.

Epidemiology (14 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions