MBA 5230 Exam 2 || with Complete Solutions.

4 Assumptions of OLS correct answers -no serial correlation -no heteroscedasticity -no multicollinearity -normality no serial correlation and test correct answers -significant relationship between a variance and itself at diff time points (repeated variable) -test with durbin-watson test. range between 0 (extremely positive) and 4 (extremely negative. 1.5-2.5 good no heteroscedasticity and test correct answers -variability of a variable is highly scattered -test with breusch-pagan. P-value 0.05 =good no multicollineraity and test correct answers -significant linear relationship among 2 or more independent variables -variance inflation factor (VIF): VIF4= good when to use robust OLS correct answers to correct undetected specification errors particularly for heteroscedasticity when to use GLS correct answers when only serial correlation found or both SC and H found Logistic Regression correct answers -use when 1 or 0 as dependent variable (dummy) -limited number of possible values, while linear is continuous ANOVA correct answers use when comparing means of 2+ groups with controlled environment R-squared correct answers goodness of model fit % of variance of the dependent variable is explained by the independent variables in the model p-value correct answers statistical significance of each independent variable. 0.001 means the independent variable is significantly related to the dependent variable when the cutoff is 0.01 coeffecient correct answers strength of relationship with dependent variable 0.1 means, when the independent variable differs by 1, the dependent variable differs by 0.1, when all other independent variables remain constant t-value correct answers relationship strength relative with dependent variable. if an independent variables absolute t-value is larger than others, it means the independent is more related to the variance of its dependent variable illustration in formula correct answers dependent variable= intercept + coefficient1* ind var 1 + coefficient2* ind var 2