B14 Regression Analysis Flashcards
What are the limitations of correlation analysis
it only mesures the strength of a linear relationship, it is strongly effected by outliers and correlation does not equate causation
Does regression try to prove causation
Yes, although there might be third factors that influence both the dependent and independent variable. When doing regression analysis we usually assume that the response variable is caused by the explanatory variable
What is a deterministic relationship in regression analysis
When the response variables value van be precisely predicted by the explanatory variables in the regression model.
What is a stochastic relationship
That the response variable is influenced by the independent variable but that there are other factors that create a difference u between the prediction and reality.
What is SSE in the OLS estimator
Sum Square Error, the linear relationship picked by OLS is the one that has the smallest Sum of Squared Errors, the errors may also be called residuals
What is the greatest limitation of the standar error of regression
It has no upper limit so it can be dificulat to determine in a vaccume.
What is the coefficient of determination
R² the share of the variance in Y that is explained by X
Which of these three measures is not a goodness of fit measure, R², SER or b1
b1
What is SST
Sum of Squares Total, aka the total variation y dependent on x. It can be broken down into explained variation and unexplained variation. SST = SSE + SSR
What is SSR
The sum square regression, aka the summed difference between what the regression model explains and what it does not you get the coefficient of determination by subtracting SSE/SST from one as it is the share of SST that is not explained by error.
Can the coefficient of determination be interpreted as the correlation of the real values with the estimated values
yes
