Brainscape's Knowledge GenomeTM

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


See full index

POTH RESEARCH > LECT 2.2: BIAS > Flashcards

LECT 2.2: BIAS Flashcards

1
Q

what is precision

A

how much uncertainty remains i the results and is related to sample size (increase ss = increase prescision = narrower CI)

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

what is bias in quantitative research

A

Bias is any systematic error AT ANY STAGE OF THE RESEARCH PROCESS (study design, data collection, analysis, or interpretation of a study that result) that prevents you from getting close to the truth and that leads to misleading or inaccurate findings/results and, thus, study conclusions

  • Deviation of results from the truth (type I or type II error, magnitude and strength of association distorted)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

explain why bias is important in terms of a type 1 error

A

Find an association b/w variables or significant finding when none exists (“spurious finding”) = type I error

=May occur by chance: 5% chance that hypothesis is not correct
=Is the improvement seen actually due to intervention?
▪e.g.: Pretest-Posttest (no control group) Pretest→ Intervention → Posttest
(Consider other possible explanations: recovery, development, other interventions received, inaccurate measure)

e.g.: Nonrandomized Control Group
Intervention 1 → Outcome
Intervention 2 → Outcome
(Consider other explanations: groups different from the start (more impaired/ chronic), different opportunities to improve (better clinic),

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

what is a type 1 error

A

Find an association b/w variables or significant finding when none exists (“spurious finding”) = type I error

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

what is a type 2 error

A

Failure to find and report an association b/w variables when a relationship truly exists = type II error
* Hypothesis is rejected, yet it is true

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

what are some reasons you might get a type 2 error

A

May be due to
* Small sample size
* Intervention not sufficiently intensive
* Quality of the measurement instrument/test
* Variation of scores on the dependent variable (individual differences b/w subjects/measurement error)

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

how is bias primarily identified

A

by critically appraising study design and methods (NOT BY LOOKING AT RESULTS USUALLY)

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

true or false: bias is primarily identified by critically approaching the resutls

A

false, the study design and methods

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

is there more bias in experimental or non experimental

A

non (observational)

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

does bias occur on a continuum or all or nothing

A

continuum (there are degrees of bias)

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

true or false: you can only see bias in the trial planning and trial implementation

A

false, also during data analysis and publication

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

explain some ways you can see “pre trial” /trial planning bias (from picture)

A

flawed study design
selection bias
channelling bias (GPT: patients are more likely to receive a specific treatment due to their characteristics, such as disease severity, comorbidities, or other factors, rather than due to random assignment)

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

explain some ways you can see bias during trial/implementation (from picture)

A

interviewier bias
recall bias
perfromance bias
missclassification of exposure or outcome etc

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

explain some ways you can see bias after trial/during analysis or publication (from picture)

A

citation bias
confounding variables

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

what ate the 3 MAIN TYPES OF BIAS

A

1) selection
2) information
3) confounding

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

“you’ve got the wrong subjects” what main type of bias

A

selection

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

“you’ve got the wrong infromation” what main type of bias

A

infromation

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

“you’ve got the wrong variables” what main type of bias

A

confounding

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

what is selection bias

A

occurs when certain groups are either excluded or overrepresented in a study (ie: YOUVE GOT THE WRONG SUBJECTS)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

when looking for selection bias, what are some things you want to look for

A

look at how the subjects were included in the study (what type of recruitment, inclusion/exclusion criteria)

look at who dropped out and the reasons (losses to follow up)

personal or disease charcaterisitcs

distribution of exposure and outcome

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

true or false: selection bias can occur at both the study design and the data analysis phase

A

true

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

what is the result of a sampling/selection bias

A

result is that the study sample is not representative of the target population

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

when is selection bias common

A

in convenience sampling (picking people that are “easy to recruit” like friends and family)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

what is the target population

A

the population about which an investigator wants to draw a conclusion

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

what is the available population

A

the population from which the sample was drawn

26
Q

what is the sample population

A

sample from a define population (should represent the target population(

27
Q

it is said that selection bias affects internal or external validity

A

external

28
Q

what is external validity

A

how applicablee the results are to the target population

29
Q

in slsection bias, we say that the study population (sample) is not representative of the target population, what are some reasons for that

A
  • Personal or disease characteristics
  • Distributionof exposure and outcome
  • Some subjects/subgroups of subjects not included in analysis (needs to be justified)
  • losses to follow-up (longitudinal study)
30
Q

true or false: selection bias effects the internal validity of a study

A

false, external(generalizability)

31
Q

what are some types of selection bias

A

volunteer bais, referral bias, non-response bias, attrition bias

32
Q

what is volunteer bias

A

study in which the participants CHOOSE if they want to be a part of the study (may be different that those who do not volunteer to be in study)

33
Q

what is referral bias

A

process of referring participants to a study or treatment introduces bias by favoring certain types of individuals or excluding others.

ex: For example, if healthcare providers selectively refer only patients with severe symptoms to participate in a study, the sample may not represent the full spectrum of the condition being studied, leading to biased conclusions

34
Q

what is non-response bias

A

Non-response bias occurs when individuals who do not respond to a survey or study invitation differ systematically from those who do respond, leading to biased results.

35
Q

what is attrition bias

A

Attrition bias occurs when participants drop out of a study over time, and their reasons for dropping out are related to the outcomes being measured

36
Q

explain information bias

A

Method of collec+ng or recording informa+on (data) regarding the study variables is flawed and yields
systematicc errors in measurement

37
Q

which type of bias affects external validity

A

sampling

38
Q

which type of bias affects internal validity

A

information
counfounding

39
Q

what are some of the ways you can get information bias

A

o Use of incorrect or non-standardized measure - Clinical Test/Measure (questionnaire)

o Measures administered or recorded incorrectly

o Incorrect classification of information (e.g. disease severity- e.g. measurement of lung function done correctly; however, patient incorrectly classified as having mild lung disease rather than severe)

o Missing or incomplete data – common in chart review (if it is random, it is not biased but if there is systematic missing information then it introduces bias)

40
Q

does information bias affect internal or external validity and explain

A

interval (accuracy)
=affects the extent to which the results accurately reflect the true relationship between the variables being studied

41
Q

what are some types of information bias

A

measurement bias, recall bias, interviews/observer/recording bias, respondent bias

42
Q

what is measurement bias

A

Measurement bias occurs when there are errors or inaccuracies in how data is collected or measured.

43
Q

what is recall bias

A

recall bias occurs when participants in a study inaccurately remember or report past events or experiences.

44
Q

what is interviewer/recording bais

A

interviewer bias occurs when interviewers or data collectors inadvertently influence participants’ responses through their behavior, tone, or body language

. Recording bias refers to errors or inaccuracies in recording data during interviews or observations,

45
Q

what is respondent bias

A

espondent bias occurs when participants in a study intentionally or unintentionally provide inaccurate or misleading responses (give socially desirable responses)

46
Q

what is a confounding variable

A

Variable or characteristic that may distort the true association between the exposure (E)/intervention (I) and outcome variable (O)

occurs when the relationship between two study variables is distorted by a third variables

47
Q

Confounding variable (“confounder”) leads to distortion in study findings/results due to variables that are :

A

o Associated with exposure (independent of outcome) AND
o Associated with outcome (independent of exposure)

48
Q

when can coufounding bias occur

A

1) if the confounding variables are not measured/not considered in the analysis

2) in group comparison, if the groups different on important chractertics that can influence the outcome

49
Q

when is confounding bias common

A

in non RCTs, between group observational designs, or single group pre-post desin

50
Q

understand the confounding examplle with stroke

A
51
Q

to address the counfounding bias, when can you minimize it (what stages)

A

design and analysis stage

52
Q

what are the 3 ways (general) to minimize confounding bias at the DESIGN STAGE

A

restriction
matching
randomizing

53
Q

explain restriction as a way to minimize bias at the DESIGN stage

A

exclude participants with certain characteristics
(ex: Restrict participant within a narrow age range (eg 18-25))

54
Q

explain matching as a way to minimize confoundingbias at the DESIGN stage

A

for each participant in one group, match with one or more participants in comparison group with same/similar value of confounding variable (e.g., match on age)
o Ensures that groups have a similar distribution of ages

55
Q

explain randomization as a way to minimize counfoundingbias at the DESIGN stage

A

random allocation to study groups balances known and unknown confounding variables between groups
o Best way to minimize confounding
o Reason why RCT is such a powerful design

56
Q

what are the 2 ways (general) to minimize confounding bias at the ANALYSIS STAGE

A

stratify
multivariate analysis

57
Q

explain stratifying as a way to minimize counfoundingbias at the ANALYSIS stage

A

can be done at design stage (straTIFIed sampling) or analysis stage (stratifed analysis)

  • Separate sample into sub-groups based on confounding variable
  • Subgroups for ages, such as 18-30, 31-40, 41-50, etc.
  • Carry out separate analysis in each sub-group
58
Q

explain multivariate anaylsys as a way to minimize bias at the ANALYSIS stage

A

analytical techniques that estimate the relationship between an exposure (or intervention) and outcome, while taking into account the effect of confounders

ex: - Relationship between stroke severity and functional mobility, independent of/separate from the effect of age
- Y=x1 +x2 +…+error (Y is outcome; x1 is exposure; x2 is confounder)

59
Q

what are the 2 ways to minimize sampling bias (general)

A

use a random sampling
use a stratified sampling

60
Q

explain how using a random sampling can minimize sampling/selection bias

A

-
Using a computer-generated random number generator.
-
Ensure all members of the population have an equal chance of being selected

61
Q

explain how using a stratified sampling can minimize sampling/selection bias

A

-
Dividing the population into subgroups (strata) and selecting a sample from each stratum
-
ensure that the sample is representative of the population as a whole.

62
Q

true or false: straitifieng the data to minimize confounding bias can only be done at the analysis stage

A

false also at the design stage

POTH RESEARCH (7 decks)

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