L4 Errors in Epi Research Flashcards

(20 cards)

1
Q

what are the two types of error?

A
  1. errors affecting precision; RANDOM ERROR

2. errors affecting accuracy; SYSTEMATIC ERROR// BIAS

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

when looking at data, what kind of error produces a deviation from the true value?

A

errors affecting accuracy, systematic errors

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

when looking at data, what kind of error produce “noise”

A

errors affecting precision, noise

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

how can random error be avoided?

A

take large samples because greater proportion of samples will be clustered around the true population mean

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

what does a confidence interval indicate

A

the likely range of the populations mean

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

what statistical approach would require you to construct a forrest plot?

A

meta-analysis

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

what are the 4 types of SYSTEMATIC ERROR

A
  1. selection bias
  2. misclassification bias
  3. confounding bias
  4. effect modification
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

what is selection bias and when does it occur

A

bias introduced to select units that are included in a study, occurs when selection procedures results in different relationship between exposure and outcome

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

how can you avoid selection bias?

A

random selection of study participants from study population ensure response rates are high and withdrawal rates are low amongst selected participants.

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

provide an example of selection bias?

A

estimating female population and taking sample from rugby match

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

what are the other kinds of selection bias? Give a brief description of them.

A

surveillance bias - looking in the wrong place for disease
referral bias - using referral clinics as a source of sample
non-response bias - refusal to participate
length of stay bias - longer stay in hospital = more likely to be asked to take part in trail
survival bias - apparent increase in prevalence of dx because we have methods of treating

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

how does misclassification bias arise?

A

errors in information that we record about our study

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

what are the four types of misclassification bias? provide brief description of each

A

recall bias - interviewees better at recalling events when disease has occurred
interviewer bias - interviewer privy to info
prevarication bias - subjects in study may have ulterior motives ie/ overestimating exposure for compensation claims
obsequiousness bias/ clever hans - subjects alter response to please interviewer

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

how can you avoid misclassification bias?

A

blind assessment of exposure and disease
use of complete and detailed sources of info
use of objective measures where available

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

describe confounding bias

A

distortion of true underlying relationship between exposure and outcome because of the influence of a third factor

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

what is a common name for confounders?

A

lurking variables

17
Q

what factor is commonly a confounding variable?

18
Q

what 3 characteristics are used to determine if a variable is confounding?

A
  1. it is causally associated with the outcome
    2 is is causally or non causally associated w/ exposure
  2. confounding variable and exposure are on TWO SEPERATE causal pathways to the outcome
19
Q

what is effect modification?

A

summary measure of association varies with the level of a second exposure variable

20
Q

______ is something you need to control for in your analyses where as _____ ______ is something you need to report.

A

a) confounding

b) Effect modification