SOAN 2120 Lecture Notes - Lecture 11: Random Number Generation, Equivalent Average, Internal Validity
Document Summary
Learning objecives: understand the concept of experimental designs using design notaion, know the various threats to internal validity, to be able to assess the impact of various threats to internal validity using the experimental design notaion. Experimental designs: independent variable -> dependent variable (diferences in outcomes between two groups, equivalent. Scores on all other variables which are likely to impact the dependent variable: diferences: manipulaion, more complicated designs. Randomly assign individuals to two or more group: random number generator, objecive: equal probability of assignment, essenial - (the groups are equivalent ). If the groups are equivalent, it is done through manipulaion. Its called selecion bias: we could compare pre-test scores (dependent variable measures) before the experiment begins. > equivalent -> proceed as if random assignment had taken place: what about all other potenial confounding variables, this happens automaically with random assignment.