EPID 301 Chapter Notes - Chapter 5: Null Hypothesis, Point Estimation, Relative Risk

• School: Queen's University
• Department: Community Health and Epidemiology
• Course: EPID 301
• Professor: Helene Ouellette-Kuntz

Document Summary

Estimates of basic parameters can be wrong because of random (stochastic) error or systematic error due to defects in study design. If 2 basic parameters are compared, they mix together (confound) the effects of several different factors. Critical appraisal: identifying vulnerabilities and weakness in research. Accept the result may be right interpret the result. Critical appraisal involves attacks (life-giving force in research, not personal attacks) The perfect study doesn"t exist (aims to identify potentially important flaws- that threatens the value of the reported information for the goals of clinical practice and public health) Random error arises when estimates of these parameters (prevalence and incidence) are inaccurate due to chance. Estimates of the parameters= educated guess about the true population parameter. Recall**: prevalence = true parameter vs. prevalence = sample-based estimate. The results would always be exactly the same which reflects that a true population parameter is not subject to random error.