Statistical Sciences 1024A/B Lecture Notes - Lecture 2: Design Of Experiments, Block Design, Confounding
Document Summary
Module 2 study design interactive video notes. Response variable the outcome of primary interest. Explanatory variable a variable that may explain or in uence change in the response variable. Confounding: two variable are confounded if their effects on the response variable are mixed up (or confused) so they cannot be distinguished. In an experiment, researchers manipulate the explanatory variable to observe its effects on the response variable. Subjects/experimental units the simulator driver participants. Levels cell phone or no cell phone: treatment (both factor and levels) speci c distraction condition (cellphone or no cellphone) Key principles of experimental design: randomization randomly assign subjects/experimental units to treatment, replication use enough subjects/experimental units, control take steps to ensure the treatments are the only differences between the treatment groups. Work together to reduce the effect of confounding. Treatment groups look similar to each other before treatments applied. helps ensure that any treatment applied is due to the treatment itself.