sources of error and bias

Question 1

error calculation

Answer 1

error=true value - estimate

Question 2

random error

Answer 2

due to sampling variation and biological variability, can be reduced with larger sample sizes

Question 3

systematic error (bias)

Answer 3

consistently over/under estimates a value, not fixed by increasing sample size, more difficult to detect or correct

Question 4

selection bias

Answer 4

caused by a non representative sample, can happen during recruitment or retention

Question 5

information bias

Answer 5

occurs during data collection,

Question 6

recall bias

Answer 6

inaccurate self reported data

Question 7

misclassification bias

Answer 7

incorrectly labeling cases and controls, non differential: affects both groups equally, differential: affects both groups differently

Question 8

confounding bias

Answer 8

occurs during analysis, a third variable influences exposure and outcome

Question 9

controlling for bias (at design stage)

Answer 9

randomization, restriction (limiting to a specific group of individuals), matching (pair cases/controls based on confounders)

Question 10

controlling for bias (analysis stage)

Answer 10

stratification: analyze subgroups separate, multivariable regression: adjusts for multiple confounders, sensitivity analysis: test how bias effects results