ECO220Y1 Lecture Notes - Lecture 21: Correlation And Dependence, F-Distribution, Linear Regression
jimmm and 37654 others unlocked
27
ECO220Y1 Full Course Notes
Verified Note
27 documents
Document Summary
Significance, and fit of a multiple regression model. A multiple regression is not a combination of simple regressions. Each coefficient tells the relationship between x and y, holding all other x variables constant. Failing to reject null for multiple regression coefficient does not mean that variable is not related to y. Means x does not contribute to model after accounting for all other x variables. Flow of foreign fighters to isis y represents number of isis fighters from a particular country: observational data, x could represent gdp per capita, muslim population, etc, 173 observations, elasticity is unit free. Changing unit of measurement does not affect coefficients. Anova: sst = (yi - )2, ssr = ( ii- )2, sse = (yi - i)2. Measure of prediction error: sst = ssr + sse, r2 = ssr/sst or 1 - (sse/sst) Fraction of variation in y explained by x.