Week 9 Flashcards
What is the difference between systematic error and random error?
Random error is due to chance such as statistical fluctuations in the measurement of data. Systematic error is a result from errors in the selection or measurement such as bias.
What is the difference between internal validity and external validity?
Internal validity is the validity of inferences drawn as they pertain to the members of the source population. External validity or generalizability is the validity of the inferences as they pertain to people outside of the source population.
When does selection bias occur in the natural history of the study?
After the choosing the population and before selecting the study participants.
When does information bias occur in the natural history of a study?
During the data collection step.
Give an example of information bias
misclassification of former smokers as non-smokers
What is non-differential misclassification?
When the misclassification of exposure status, for instance, is equal in the diseased and non-diseased groups (ie, it does not differ by disease).
How does non-differential misclassification, which is independent of disease status, affect the estimates in a study?
non-differential misclassification biases toward the null (no effect)
Twenty percent of the nonusers with CHD were categorized as users. Everyone else was classified accurately according to the truth. Is this an example of misclassification of exposure or disease?
misclassification of exposure
Twenty percent of the nonusers with CHD were categorized as users. Everyone else was classified accurately according to the Truth. What type of misclassification is this an example of?
Differential misclassification of exposure, because the misclassification of exposure status is dependent on disease status.
How can differential misclassification affect the estimates of study?
Differential misclassification can bias either toward the null or away from the nul.
To have internal validity, a study must successfully address selection bias, information bias, and confounding, but not
generalizability
In an observation study, threats to external validity include
effect modification
What is validity?
The degree to which the inferences drawn from a study are warranted when account is taken of: the study methods, the representativeness of the study sample, and the nature of the population from which it is drawn.
What is internal validity?
the credibility with which statements can be made about whether the association between the exposure and disease found in your study is a true association; what you observe is not due to random error, confounding or bias
What is external validity (generalizability)?
validity of the inferences as they pertain to other circumstances; ability to generalize the results from a given study to populations beyond the study subjects.
What study design is limited in generalizability?
clinical design, experimental study design; however, the bigger the sample, the more it is generalizable
What are some threats to internal validity?
chance (random error), confounding, bias
What are some threats to external validity?
internal validity, choice of study population, effect modification
Why is effect modification a threat to external validity?
if the exposure causes disease only in the presence or absence of some third variable, generalizability may be limited
What is bias?
the result of a systematic error in the design or conduct of a study; observed study results will tend to be different from the true results; happens in nature, unlike confounding, relates to the process of how the study is conducted
bias can lead to incorrect:
measures of disease occurrence or association, estimate of statistical significance
what is selection bias?
individuals have different probabilities of being included (selected, recruited, retain) in the study sample according to relevant study characteristics (exposure/disease)
What is information bias?
systematic tendency for individuals selected for inclusion in the study to be erroneously placed in different exposure/disease categories. (ex: misclassification)
When can selection bias occur in cohort studies?
if the frequency of risk factors in the exposed group(s) is different than that in the comparison group; more commonly, differential losses to follow-up
What is control selection bias in case control studies?
when the selection of controls in case-control study is associated with exposure status
What is Berksonian Bias?
occurs when cases and controls are selected from a subpopulation instead of the general population; when both cases and controls are sampled from a hospital
information bias examples
recall bias, measurement bias, misclassification, interviewer bias, observer bias, respondent bias
how can we minimize information bias?
getting expert opinion, using validated instruments and objective measures, standardized protocols and training data collection team
what is interviewer bias?
behavior of the interviewer may influence respondent answers
In what study designs are exposure-related biases more likely to be present?
case-control (population based)
what is observerr bias?
ascertainment of outcome is influenced by knowledge of exposure or outcome status
In what study designs are outcome-related biases more likely to be present?
cohort
what is non-differential misclassification of disease?
non-differential misclassification of disease; disease is misclassified proportionally among exposed and non-exposed
what is non-differential misclassification of disease?
non-differential misclassification of disease; disease is misclassified proportionally among exposed and non-exposed
What is non-differential misclassification of exposure?
exposure is misclassified proportionally among diseased and non-diseased
How can one measure the ability to distinguish between who does and does not have a disease (validity/truth)?
sensitivity and specificity
what is sensitivity?
the ability of a test to correctly identify those who have the characteristic of interest (a/a+c)
what is specificity?
the ability of a test to correctly identify those who do not have the characteristic of interest. (d/d+b)
what type of bias is lead time bias?
information bias
what type of bias is length bias?
selection bias
what is lead-time bias
overestimation of survival duration due to earlier detection due to earlier detection by screening than clinical presentation
what is length bias?
occurs when some people who are screened have disease that progresses more slowly than all people who have the disease.
what are examples of cross-sectional biases?
incidence-prevalence bias, temporal bias
what is incidence-prevalence bias?
including prevalent cases in a study when the goal is to make inferences about incident cases
what is temporal bias?
cannot establish temporal sequence between exposure and outcome
observer bias and respondent bias are common in cohort designs. true or false?
true
recall bias and interviewer bias are common in case-control designs. true or false?
true
