Trans - Biases in Research Flashcards
validity
closeness to the truth
reliability
closeness of the observations to each other
random error
systematic deviations away from the truth (noise)
systematic error
a process at any stage of inference tending to produce results that depart systematically from the true values
selection bias
occurs when a selection procedure favors one treatment over another
selection bias reduced by:
sampling
adherence to inclusion and exclusion criteria
random allocation of treatment
performance bias
occurs when a study procedure favors one treatment group over another
performance bias reduced by
blinding
standardization of procedures
detection bias
occurs when outcome assessment favors one group over another
detection bias reduced by
blinding
standardization of tools and procedures
pretesting
exclusion bias
occurs when a procedure removes a subject or treatment group from data analysis
confounding bias
occurs when a risk factor is closely associated with the one of interest
the effect of one is confused with or distorted by the effect of the other
reporting bias
occurs when part of the research findings are not reported
the process of deciding which subjects should get what treatment
allocation
process of subdividing the subject group into homoegenous subgroups before or after allocation
stratification
good allocation
- blinded
- future assignments cannot be predicted from past assignments
- reproducible
- documented
process of allocating the subject to the treatment by an auditable random process
randomization
performance of additional diagnostic or therapeutic acts on the experimental group but not on the control
co interventions
co interventions result in:
spurious increase in the difference of clinical outcomes observed
control patients accidentally receive treatment
contamination
contamination results in:
spurious decrease in the difference of the clinical outcomes observed
noncompliers
patients who do not drop out but do not follow the treatment
withdrawals
patients who decline to continue or are dropped from the study
drop outs
results will not be known
