STS 350 Study Guide - Midterm Guide: Ordinary Least Squares, Ordinary Differential Equation, Confidence Interval

• School: Drexel University
• Department: Statistics
• Course: STS 350
• Professor: Michael Bruneau

Document Summary

Discussion board 3: ordinary least squares regression: relationship between independent and dependent variable. Independent variable: variable that is expected to influence the dependent variable (explanatory variable!!: dichotomous variable: can only take on 2 possible variables, ordinal variable: any 1-unit difference is the same difference in the amount of property being measured. Interval variable: things that are hard to quantify (like good, bad, awful) hard to define the scale: dependent variable: phenomenon explained by some variable (independent) Linearity: looks at the relationship between iv & dv looks at which variables are good predictors of the dv: multicollinearity: when independent variables are related to one another. Measure of fit (r2): between 0-1 the proportio(cid:374) of (cid:448)aria(cid:374)(cid:272)e i(cid:374) the dependent variable that"s e(cid:454)plai(cid:374)ed (cid:271)(cid:455) the i(cid:374)depe(cid:374)de(cid:374)t (cid:448)aria(cid:271)les: 1 = perfect fit (more explanation, 0 = bad fit (less explanation) If it"s (cid:271)igger tha(cid:374) +-1. 96 the(cid:374) it"s statisti(cid:272)all(cid:455) sig(cid:374)ifi(cid:272)a(cid:374)t (cid:894)if that"s (cid:455)our preset value) Regressio(cid:374): (cid:862)(cid:449)hat is the effe(cid:272)t of (cid:894)differe(cid:374)t(cid:895) (cid:894)le(cid:448)els of) independent.