Exam 4 Study Guide Flashcards
What results from a systematic error in ascertaining information from study subjects?
Information bias
What results from a systematic error in selecting/recruiting study subjects?
Selection Bias
What is a distortion that results from procedures used to select study participants and/or that influence participation in the study?
Selection bias
Study subjects who are recruited solely from clinical population may have more diseases and adverse health conditions than the general population describes what?
Selection bias
Volunteers may differ from the general population with regard to both disease and exposure status describes what?
Selection bias?
Subjects who withdraw from long term cohort studies or who are “lost to follow up” May differ from other cohort members with respect to exposure and disease status describes what?
Selection bias
Error in the measurement of exposure, disease, and/or other key variables which can result in misclassification of study subjects with regard to exposure, disease, and/or other key variables describes what?
Information bias
Degree of misclassification varies among the study groups ex: exposure is misclassified among cases but not controls. Can either enlarge or diminish the observed association between exposure and outcome describes what?
Differential misclassification
Misclassification is randomly distributed among the key study groups and almost always diminishes the magnitude of the association between exposure and outcome describes what?
Non-differential misclassification
Loss to follow up in a cohort study is a form of what kind of bias?
Selection bias
What is the distortion of the association between an exposure and a health outcome by a third variable that is related to both?
Confounding
What must be associated with the exposure and be independently associated with the outcome, cannot be in the causal pathway, and will result in an overestimate or underestimate of the true association between an exposure and an outcome?
Confounding variable
What are 3 ways to overcome confounding in the design of the study?
Matching, restriction, randomization
What are 2 ways to overcome confounding in the analysis?
Stratification (restriction), statistical adjustment (multivariate modeling, multivariate adjustment)
What refers to the degree to which your tools of choice are measuring what they are designed to measure?
Validity
What deals with measurement precision/consistency and reflects the degree to which measures are free from random error?
Reliability
What are 5 examples of observational studies?
Descriptive, ecological, cross-sectional, case-control, cohort
What are 4 examples of interventional studies?
Randomized control trials, clinical trials, field trials, community trials
What is the key purpose of of random assignment in a randomized control trial?
Equalize the effects of extraneous variables among the study groups
What is the purpose of a double blind in a randomized control trial?
Minimize observer and subject bias
