Statistical techniques which reveal (or not) relationships among variables are useful in law, business, and government, as well as in academic disciplines, particularly social sciences. These techniques have immediate application in senior theses and project papers. Examples: determinants of voting behavior; influences on executive compensation; crop yields and emigration in 19th C. Europe. This minicourse will introduce CSS juniors and seniors to some basic tools, including the development of hypotheses, equation specification, and simple and multiple regression, with associated statistics (coefficient estimation, t-statistics, statistical significance, and R squared) and with attention to several problems (multicollinearity, omitted variables). Likely class time: three or four sessions plus several lab sessions, to allow students to experience the frustration, excitement, and satisfacation of empirical investigations. No calculus required. A CSS elective course. |