MARK312 Lecture Notes - Lecture 8: Correspondence Analysis, Observational Error, Crystal Computing
Document Summary
Evidence of causality to demonstrate causation, a researcher must be able to show three things: concomitant variation (correlation) - a statistical relationship between two variables. A and b must vary together, either positive or inverse: appropriate time order of occurrence - a change in an independent variable occurring before an observed change in the dependent variable. A must occur before b: elimination of other possible causal factors - need to make sure that change in b was not caused by some other factor than a. Terms experimental design - the test in which the researcher las control over and manipulates one or more independent variable. Designing an experiment the treatment, or experimental, variable - the independent variable that is manipulated the participants/ subjects - who are part of the experiment the experimental design includes four factors: Day after recall test percentages of viewers is compared to historic norms. Have you seen any ads in the last day - un-aided.