HLTH333 Lecture Notes - Lecture 8: Measuring Instrument, Recall Bias, Social Desirability Bias
Document Summary
Random error: reduces the precision of the estimates (or, rr, etc) Or or rr will be smaller than the true or or rr. Systematic error: reduces the validity or accuracy of the estimates. Results are similar to one another, but the scores do not represent the score of the target population. Neither accurate of precise = systematic and random errors. Study of cancer patients and coffee consumption - study showed that those who drank less coffee has less incidence of cancer. The study was biased because the control drank less coffee than the general population. The control had unusually low levels of exposure, which led to an increased measure of association in the study. Example of systematic error - systematic recruitment of controls was flawed. Internal validity: degree to which a study provides unbiased estimates of what it claims to measure. The extent to which the study result mirrors the truth in the study population.