COGS 14A Lecture Notes - Lecture 19: Experiment, Infant Mortality, Time Series
Document Summary
Cogs 14a lecture 19: quasi experiments continued 3/1/17. Would be ideal to have true experiments all the time, however sometimes the logistics of the experiment prevent that. Sometimes you can"t corm proper control groups or limit confounds. A group with no randomization, use a pretest and a posttest to check if an event had any impact on the dependent measure. Allow you to gather important data, but without randomization required in experimental studies. These will have a control group or multiple measures (or both) These are stronger than pre-experimental designs, but will have weaker casual explanations than true experiments. Allow you to compare the result of a manipulation of the independent variable with a different group of people the control group should be similar but assignment to the treatment and control groups is not random. Suppose we compare mortality rate with a comparable state that does not institute a prenatal program.