Flashcards
ECON375 Dorian
Cards in this deck (24)
- 1What are the 5 SLR assumptions1) Linear in parameters
2) Random sampling
3) Sample variation in xs
4) Zero conditional mean E(u|x)=0
5) Constant variance Var(u|x)=σ² - 2What assumptions need to hold for an estimator to be unbiased?SLR 1-4
- 3What is BLUE?Best Linear Unbiased Estimator
- 4What is SSR?Sum of Squared Residuals, sum of u hat squared.
- 5What is R squared?R squared is a measure of goodness of fit; how much explanatory power our model has. It is SSE/SST or (1-SSR)/SST, where
SSE = explained sum of squares and
SST = total sum of squares - 6What is the standard matrix form?y = Xβ+u
- 7When can matrices be added?When they have the same dimensions
- 8When can matrices be multiplied?When they conform; A = (m*n), B = (n*p), the ns make them conformable, and will create the matrices AB = (m*p)
- 9Multiply the following matrices: 2, -1, 0 0, 1, 6, 0 -4, 1, 0 * -1, 2, 0, 1 3, 0, 0, 02*0 + -1*-1 + 3*0 = 1
2*1+ -1*2 + 0*0 = 0
2*6 + -1*0 + 0*0 = 12
2*0+ -1*1 + 0*0 = -1 etc.
1, 0, 12, -1
-1, -2, -24, 1 - 10What is an identity matrix, and what does it mean?An identity matrix is (X'X)⎺¹X'X, and it is essentially a 0 matrix.
- 11What does unbiasedness mean?Although an OLS estimator does not get the true population mean in every sample, it gets a mean close to the population mean over a range of samples.
- 12What is the Gauss Markov Theorem?If SLRs hold, OLS is BLUE
- 13What does asymptotically consistent mean?As n → ∞, βhat = β₁. This is also known as plim.
- 14If an estimator is unbiased, is it consistent?Yes
- 15If an estimator is consistent, is it unbiased?Not necessarily. We can show examples with biased estimators that regardless collapse on the true value of β₁
- 16What is the CLT?Central limit theorem - as n gets large, we can assume normality
- 17What is a p value?The probability of observing a test statistic at least as extreme as the one observed in the sample, given H₀ is true
- 18What is the difference between a one and two sided test?A one sided test looks for the 5% level of significance in 1 tail. On stata, we would therefore observe the 10% level, but one side of it.
- 19What is a Type I error?Rejecting the null, given that it is true
- 20What is a Type II error?Failing to reject the null, given that it is false
- 21What is the power of a test?1 - P(Type II error)
- 22What does a 95% confidence interval mean?We can be sure that 95% of intervals calculated in this manner will contain the true value β₁.
- 23What are the 6 MLRs?1) Linear in parameters - y = Xβ + u
2) Random sampling
3) Matrix X has full column rank [(X'X)⎺¹ exists, no exact linear relationship between regressors]
4) Zero conditional mean
5) Homoskedastic random errors
6) Normality - 24What is the partialling out interpretation?We can perform multiple regression as a series of simple regressions.
1) regress S on EXP, save resids as ES. ES represents S purged of the influence of EXP
2) purge Earn of the influence of EXP, save resids as EE.
3) Regress EE on ES.
Information
- Description
- Mathematics Flashcards on ECON375 Dorian, created by Matt Ditchburn on 20/10/2018.
- Created by
- Matt Ditchburn
- Views
- 1
- Rating