MATH14500 Lecture Notes - Lecture 2: Confounding, Treatment And Control Groups
14 Sep 2016
School
Department
Course
Professor
Document Summary
Observations- qualitative studies where opinions or cultural norms or settings are explored by immersion or by interviewing. Experiments- usually quantitative study where a treatment is imposed in order to measure a possible change in one or more variables. Association- when values of one variable tend to be related to another. Causation- when a treatment of one variable changes the values of another. Change the explanatory to see if the response changes. Give the class coffee at 10 pm and the other none. See how alert the class is the next morning. We can only assume causation if we take a random sample. Control group (placebo has to be like treatment sometimes results occur like treatment) (eliminates bias) *must be a randomly assigned process for about equal sized groups. One other typical type of an experiments matched pairs - we examine the difference between treatments on each case.
