Bias, Confounding and Interaction

Question 1

What are some examples of random error?

Answer 1

Biological variation and sampling errors
- they are usually evenly distributed above and below the true value

Question 2

What are systematic errors also known as?

Answer 2

Also known as bias (any deviation that is not due to chance alone)

Question 3

What is the definition of precision?

Answer 3

Precision refers to how well repeated observations agree with each other

Question 4

What is the definition of accuracy?

Answer 4

Accuracy refers to how well the observed value agrees with the true value

Question 5

What are some examples of systematic error?

Answer 5

Selection bias, Information bias, Confounding bias

Question 6

What are the two different types of information bias

Answer 6

Recall bias and Observer bias

Question 7

What is the definition of bias?

Answer 7

a systematic deviation of results or interferences from the truth or any processes that lead to systematic deviation

Question 8

What is the definition of selection bias?

Answer 8

systematic difference in the enrolment of participants in a study
that leads to an incorrect result (e.g., risk ratio or odds ratio) or inference.

Question 9

What is the definition of information bias?

Answer 9

systematic difference in the collection of data regarding the participants in a study
e.g
* incomplete/missed medical records
* incomplete questionnaires

Question 10

What is the definition of error?

Answer 10

The difference between the true value and the observed/ estimated one

Question 11

What is the definition of random error?

Answer 11

Unknown and unnpredictable changes in the experiment of observation

Question 12

What are some major sources of random error?

Answer 12

• Biological variation
• sampling errors
• measurement errors

these are all evenly distributed above and below the true value

Question 13

Where do systematic errors have a tendency to be?

Answer 13

Either above or below the true value

Question 14

What is the net random error over a large number of observations?

Answer 14

around zero, estimate may be imprecise but not inaccurate

Question 15

What are measurement errors referred as?

Answer 15

misclassification

Question 16

What is the definition of information bias/ misclassification?

Answer 16

• Incomplete medical records
• Patients completing questionnaires incorrectly
• misinterpretation of records

Question 17

What are the two different types of information bias?

Answer 17

• Recall bias- differential recall of information by cases and controls
• Observer bias- bias in information collected by the investigator

Question 18

What is an example of observer bias?

Answer 18

Those affected by a disease may make more of an effort to remember details

Question 19

What is an example of recall bias?

Answer 19

More likely to remember the overly heavy/ light foetuses that were aborted

Question 20

What is the definition of confounding?

Answer 20

The distortion of the association between an exposure and a health outcome by a third variable that is associated to both

Question 21

What must a variable be to be a ‘counfounder’?

Answer 21

• The factor must be independently associated with the disease
• The factor must also be associated with the exposure being investigated
• It should not lie on the causal pathway between exposure and disease

Question 22

What is the effect of confounding?

Answer 22

• Confounding factors can lead to bias in the estimate of the impact of the exposure being studied

Question 23

What are four ways you can deal with confounding?

In the design stage preferably

Answer 23

• Restriction- only select units in relation to the cofounder
• Randomisation- clinical trials so there is a similiar distribution of the cofounder in both groups
• pair matching- selecting for each case, one or more controls with similiar characteristics
• frequency matching- ensuring that as a group the cases have similiar characteristics to the controls

Question 24

How can you deal with confounding at the analytical stage?

Answer 24

• Stratification- estimate measures of association within different levels of the confounding factor
• Statistical modelling- multivariate analyses can control for confounding

Question 25

What is the definition of the odds ratio?

Answer 25

The odds ratio represents the odds that an outcome will occur in a particular exposure, compared to the odds of the outcome occuring in the absence of that exposure

Question 26

What does an odds ratio of 1 >1 and 1< mean

Answer 26

1= esxposure does not effect the odds
>1 = exposure associated with higher odds
<1= exposure associated with lower odds

Question 27

How would you know that something is a confounding variable in terms of odds ratio?

Answer 27

If you take it out as a variable and there is a difference in the odds between that and the crude odds ratio

Question 28

How will you know if there is effect modification based on the stratum-specific odds ratio?

Answer 28

The startum-species estimates will differ from eachother entirely

Question 29

Summarise confounding

Answer 29

Remove or account for the effect of the confounder in order to get nearer to the truth (actual measurement of effect).
Effect of the exposure is the same for all categories of the confounding factor

Question 30

Summarise interaction

Answer 30

Interaction is an important property of the association between two factors (detect and describe)
The factor modifies the effect of other: effect varies across category of the other factor