STATS 13 Chapter Notes - Chapter 4.2: Confounding, Size 14, Scientific Control
Document Summary
Scientific studies involve a search for cause-and-effect relationships between variables. The main challenge in designing a good study is to avoid the effect of confounding variables. To deal with unseen confounders, we rely on a strategy called randomization or random assignment. The importance of random assignment to avoid confounding effects leads to one of the most important distinction in all of science: the difference between an observational study and an experimental study. Observational: the values of the explanatory variable are simply observed. Experiment: the investigator get to assign the conditions to the different subjects. Study: 48 graduate students in business participated in a bargaining game. Researchers kept track of whether or not the students misrepresented (lies about) the size of the pot when they were negotiating with other players. Step 2: design a study and collect data. Response variable: whether or not the person lied. Suppose there were 30 male and 18 female subject s.
