POL SCI 375 Study Guide - Null Character, Statistical Hypothesis Testing, Statistical Significance
Document Summary
For all hypothesis tests: we choose a null hypothesis to disprove, usually the opposite of what we want to demonstrate (our alternative hypothesis), we accept some small chance that we will draw the wrong conclusion and reject the null hypothesis when it is true (make a type 1 error), usually a 5% chance, we conduct a statistical test to show that, if the null hypothesis is true, our results would be extremely unlikely. Minimizing errors : specifically, the regression line minimizes the squared distances between the predicted values and the observed values, we call it the ordinary least squares line or the ols regression line, it gives the lowest possible sum of the squares of errors in prediction, these prediction errors are called residuals, the regression line has the lowest possible sum of squared distances to the points (residuals) by using square law .