Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
36 Cards in this Set
- Front
- Back
Types of data |
Time series Cross sectional Pooled |
|
What is another name for qualitative variables |
Dummy variables |
|
What is panel data? |
Same Cross sectional data surveyed over time. |
|
What is non eperimantal data? |
Data controlling agency does not have direct control over the data. Same data compatible with more than one theory. Passive data. |
|
What is a parameter? |
Numerical quantity that characterised a population or some aspect of it. Also know as regression co-efficients. |
|
What is the scope of regression analysis? |
To explain the relation of the dependent and independent variables allowing for an inexact relationship which is represented by the error term. Estimate the PRF. |
|
What are residuals? |
The difference between the predicted and the observed values due to errors. |
|
Multiple linear regression |
A regression with two or more independent variables. |
|
Define income and substitution effects |
Income effect = effect of increased purchasing power on consumption. Substitution effect = how consumption changes with relative changes in income and price. |
|
What is the conditional mean? |
Mean dependent variable value for a particular value of the independent variable. |
|
Population regression line |
Line adjoining all conditional means of all independent variables. |
|
Population regression function |
Mathematical form in which the population regression function is expressed. |
|
Elaborate on the intercept |
The intercept is the conditional mean of dependent variable if independent variables are zero. |
|
What is the significance of the slope |
The slope is the rate of change of the conditional mean in response to an independent variable. |
|
What is a random variable |
Numerical description of the outcome of a statistical experiment which is characterised by a porbability distribution. |
|
Discrete vs continuous random variables? |
Discrete random variables can only have a finite number on the number line Continuous random variables also includes any value in the interval between finite numbers |
|
4 Reasons for error term |
1. Influence of variables that are not explicitly included in the model. 2. Intrinsic randomness 3. Error of measurement 4. Ockham's razor |
|
Why isn't the sample regression line equal to the population regression line? |
1. Sampling fluctuations 2. Sampling errors |
|
What is the residual sum of squares |
The sum of all the errors of the same squared |
|
How do you estimate b1 and b2 |
By solving the least squares nornal equations simultaneously |
|
What is the equation or the SRF and why? |
Sample mean of Y = b1 + b2( sample mean of X) |
|
What is Okun's law? |
Yt = -0.4(Xt - 2.5), relationship between economic growth and unemployment. |
|
Percentage point increase vs percentage increase |
Percentage point increase = increase by 1% Percentage increase = (7 - 6) / 6 = 16.6% |
|
Name Gauss Markov assumptions |
1. Linear parameters 2. Given Xi the expected value of ū = 0 therefore zero conditional mean 3. There is random sampling and sample variation in x 4. Homoscedasticity 5. No autocorrelation between error terms 6. No correlations between dependent and independent variables (exogeniety) 7. No correlation between independent variable and error term. |
|
Homoscedasticity vs heteroscedasticity |
Homoscedasticity - size of the error term is the same in all the independent variables Heteroscedasticity - size of the error term varies with every independent variable. |
|
What is autocorrelation? |
The covariance between variables is 0 and they do not effect each other. |
|
Define positive and negative autocorrelations |
Postive autocorrelation - variable above mean will influence another variable to be above it.
Negative autocorrelation - value below mean will influence another variable to be below it. |
|
What is Okun's law? |
For every 1% increase above 2.5% GDP, there is a 0.4% decrease in unemployment. |
|
What are two other names for the independent variable |
Explanatory variable, regressor |
|
What are the assumptions of the classical linear regression called? |
The Gauss Markov theorem, OLS estimators are BLUE |
|
What is a BLUE? |
Best linear unbiased estimator, Ordinary least squares, lower variance. If biased there is no sense in comparing their estimates. |
|
When deriving OLS estimators what are important things to note? |
1. Differentiation of b0 and b1 both = 0 2. Whole digits multiplied by given equation can be cancelled with 0 3. Separate sums 4. Remove averages from the sums 5. Turn average of x into xi |
|
What is the law of iterated expectations? |
Expectation of the estimate of B1 is infact B1 and the estimator is unbiased can be proved by proving thats expected B1 estimate given x is B1 |
|
What is the optional Gauss Markov assumption and why? |
The optional assumption is that the error follows a normal distribution so that confidence intervals may be calculated. |
|
What is multicollinearity? |
When independent variables of a regression model are correlated. |
|
What is am MLE |
Maximum likelihood estimator |