SOAN 2120 Chapter Notes - Chapter 7: Random Assignment, Dependent And Independent Variables, Regression Analysis
Document Summary
Valid comparisons depend on comparing things that are fundamentally alike. To be compared two groups must e similar in most respects. Random assignment: method for assigning cases (e. g. individuals, organizations, etc) to groups for the purpose of making comparisons. Each person has exactly equal chance of ending up in one group or the other. In the hopes that the two groups, in the end, will be as similar as possible represent that population. Vs. random sampling in a survey: have a population and randomly pick a sample group to treatment or independent variable (what the researcher modi es) In research reports researchers give names of designs instead of describing them. Standard designs illustrate common ways to combine design parts. All designs are variations of classical experimental design (random assignment, pretest and posttest, experimental group, control group) Preexperimental designs: lack random assignment and are compromises or shortcuts. Used in situations where it is dif cult to use classical design.