Brainscape's Knowledge GenomeTM


Top Flashcards (Certified)
All Flashcards

POTH RESEARCH > LECT 2.2: BIAS > Flashcards

LECT 2.2: BIAS Flashcards

(62 cards)

Start Studying
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
what is the available population
the population from which the sample was drawn
26
what is the sample population
sample from a define population (should represent the target population(
27
it is said that selection bias affects internal or external validity
external
28
what is external validity
how applicablee the results are to the target population
29
in slsection bias, we say that the study population (sample) is not representative of the target population, what are some reasons for that
- 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
true or false: selection bias effects the internal validity of a study
false, external(generalizability)
31
what are some types of selection bias
volunteer bais, referral bias, non-response bias, attrition bias
32
what is volunteer bias
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
what is referral bias
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
what is non-response bias
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
what is attrition bias
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
explain information bias
Method of collec+ng or recording informa+on (data) regarding the study variables is flawed and yields systematicc errors in measurement
37
which type of bias affects external validity
sampling
38
which type of bias affects internal validity
information counfounding
39
what are some of the ways you can get information bias
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
does information bias affect internal or external validity and explain
interval (accuracy) =affects the extent to which the results accurately reflect the true relationship between the variables being studied
41
what are some types of information bias
measurement bias, recall bias, interviews/observer/recording bias, respondent bias
42
what is measurement bias
Measurement bias occurs when there are errors or inaccuracies in how data is collected or measured.
43
what is recall bias
recall bias occurs when participants in a study inaccurately remember or report past events or experiences.
44
what is interviewer/recording bais
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
what is respondent bias
espondent bias occurs when participants in a study intentionally or unintentionally provide inaccurate or misleading responses (give socially desirable responses)
46
what is a confounding variable
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
Confounding variable (“confounder”) leads to distortion in study findings/results due to variables that are :
o Associated with exposure (independent of outcome) AND o Associated with outcome (independent of exposure)
48
when can coufounding bias occur
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
when is confounding bias common
in non RCTs, between group observational designs, or single group pre-post desin
50
understand the confounding examplle with stroke
51
to address the counfounding bias, when can you minimize it (what stages)
design and analysis stage
52
what are the 3 ways (general) to minimize confounding bias at the DESIGN STAGE
restriction matching randomizing
53
explain restriction as a way to minimize bias at the DESIGN stage
exclude participants with certain characteristics (ex: Restrict participant within a narrow age range (eg 18-25))
54
explain matching as a way to minimize confoundingbias at the DESIGN stage
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
explain randomization as a way to minimize counfoundingbias at the DESIGN stage
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
what are the 2 ways (general) to minimize confounding bias at the ANALYSIS STAGE
stratify multivariate analysis
57
explain stratifying as a way to minimize counfoundingbias at the ANALYSIS stage
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
explain multivariate anaylsys as a way to minimize bias at the ANALYSIS stage
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
what are the 2 ways to minimize sampling bias (general)
use a random sampling use a stratified sampling
60
explain how using a random sampling can minimize sampling/selection bias
- Using a computer-generated random number generator. - Ensure all members of the population have an equal chance of being selected
61
explain how using a stratified sampling can minimize sampling/selection bias
- 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
true or false: straitifieng the data to minimize confounding bias can only be done at the analysis stage
false also at the design stage

POTH RESEARCH (7 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2025 Bold Learning Solutions. Terms and Conditions