econometrics 1 flashcards

0.0(0)
Studied by 0 people
call kaiCall Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/100

encourage image

There's no tags or description

Looks like no tags are added yet.

101 Terms

1
New cards

random variable

a function that maps random outcomes to real numbers


  • when X is discrete = discrete variable

  • when X is interval = continuous random variable


2
New cards

cumulative distribution function

gives the probability, for any real number x, that X is below x

<p>gives the probability, for any real number x, that X is below x</p>
3
New cards

probability distribution of discrete variable X

function that, for any real number x, gives the probability that X=x

4
New cards

probability distribution function (density) for continuous variables

denoted as f(x) and is the derivative of cdf F(x) at x

<p>denoted as <em>f(x)</em> and is the derivative of cdf F(x) at x</p>
5
New cards

expected value (expectation/ mean)

E(x) of random variable X is the long-run average of X over many repeated trials

  • weighted average of all possible values of X


<p>E(x) of random variable X is the long-run average of X over many repeated trials</p><ul><li><p>weighted average of all possible values of X</p></li></ul><p></p>
6
New cards

expectation is a linear operator

E(aX +bY) = aE(X) +bE(Y)

7
New cards

variance V(X)

the expected (squared) distance of X from the mean


<p>the expected (squared) distance of X from the mean</p><p></p>
8
New cards

variance is a quadratic operator

if a is a constant,

V(a)=0 V(aX)= a²V(X)

9
New cards

standard error

is the squared root of the variance


<p>is the squared root of the variance</p><p></p>
10
New cards

joint probability distribution of (X,Y)

characterised by the probabilities Pr(X=x, Y=y) for all possible combinations of values (x,y)

eg. P(Y=1)=P(X=0, Y=1) + P(X=1,Y=1)

<p>characterised by the probabilities Pr(X=x, Y=y) for all possible combinations of values (x,y)</p><p>eg. P(Y=1)=P(X=0, Y=1) + P(X=1,Y=1)</p>
11
New cards

conditional distribution

conditional distribution of Y given X=x is defined by:


<p>conditional distribution of Y given X=x is defined by:</p><p></p>
12
New cards

bayes rule

knowt flashcard image
13
New cards

conditional expectation

the conditional expectation of Y given that X=x is

<p>the conditional expectation of Y given that X=x is</p>
14
New cards

conditional variance

the conditional variance of Y given X=x is:

<p>the conditional variance of Y given X=x is:</p>
15
New cards

law of iterated expectations (proof included)

for 2 random variables X,Y

E(Y) = Ex [ E(Y|X) ]

→ also works when conditioning on more than one variable E(Y)=E[ (Y| X,Z) ]

<p>for 2 random variables X,Y</p><p><strong>E(Y) = Ex [ E(Y|X) ]</strong></p><p><strong>→ </strong>also works when conditioning on more than one variable E(Y)=E[ (Y| X,Z) ]</p>
16
New cards

covariance ( and properties)

cov (X, Y)=cov (Y,X)

cov (a, X) = 0

cov (bX, Y) = b . cov(X, Y)

cov (X+Z, Y) = cov (X,Y) + cov(Z,Y)

cov( a+bX, c +dY) = b. d. cov(X, Y)

<p>cov (X, Y)=cov (Y,X)</p><p>cov (a, X) = 0</p><p>cov (bX, Y) = b . cov(X, Y)</p><p>cov (X+Z, Y) = cov (X,Y) + cov(Z,Y)</p><p>cov( a+bX, c +dY) = b. d. cov(X, Y)</p>
17
New cards

variance of a sum X+Y

V(X+Y)= V(X) +V(Y) +2cov(X,Y)

18
New cards

correlation

the correlation of X and Y can be defined by:


<p>the correlation of X and Y can be defined by:</p><p></p>
19
New cards

independence

two variables X and Y are independent when knowing the value of X foes not provide any information on Y. distribution of one variable doesn’t depend on realisations of the other variable

<p>two variables X and Y are independent when knowing the value of X foes not provide any information on Y. distribution of one variable doesn’t depend on realisations of the other variable</p>
20
New cards

if X and Y are independent then:

pr(X=x, Y=y) = Pr(X=x) . Pr(Y=y)

E(X . Y)=E(X) . E(Y)

cov(X, Y) = corr(X, Y) = 0

V(X+Y)=V(X) + V(Y)

21
New cards

a continuous random variable X follows a normal distribution( gaussian distribution) when its density is:

where μ is the expectation, and σ² its variance

usually denoted as N (μ, σ²)

<p>where μ is the expectation, and σ² its variance</p><p>usually denoted as <em>N</em> (μ, σ²)</p>
22
New cards

the normal distribution with zero mean and unit variance is denoted by:

N (0, 1)

its density is usually denoted as ϕ

<p><em>N</em> (0, 1)</p><p>its density is usually denoted as ϕ</p>
23
New cards

normal distribution is symmetric around its mean

for Z normal with zero mean : p(Z > c) = p(Z < -c)

for the standard normal: 1-ϕ (c ) = ϕ (-c )

<p>for Z normal with zero mean : p(Z &gt; c) = p(Z &lt; -c)</p><p>for the standard normal: 1-ϕ (c ) = ϕ (-c )</p>
24
New cards

sample

is a finite collection of n observations of Y : {Y1, …, Yn}

→ is independently and identically distributed if:

  • for any i, Yi is drawn at random from the distribution of Y

  • for any i, j where i≠j, Yi and Yj are independent


25
New cards

sample average (mean) + properties

the sample average of a given iid sample is denoted as Ybar and given by

  • if Y is normal and the sample is iid, then Ybar n is also normal → as it is the sum of independent normal variables

  • Ybar n and its distribution can be approximated when n is relatively large using law of large numbers & central limit theorem


<p>the sample average of a given iid sample is denoted as Ybar and given by</p><ul><li><p>if Y is normal and the sample is iid, then Ybar n is also normal → as it is the sum of independent normal variables</p></li><li><p>Ybar n and its distribution can be approximated when n is relatively large using <em>law of large numbers &amp; central limit theorem</em></p></li></ul><p></p>
26
New cards

the expectation and variance of the sample average

knowt flashcard image
27
New cards

law of large numbers

in an iid sample, Ybar n converges in probability to E(Y)

→ for any c>0, the probability that the distance between Ybar and E(Y) is larger than c, goes to 0 and n goes to infinity

<p>in an iid sample, Ybar n converges in probability to E(Y)</p><p>→ for any c&gt;0, the probability that the distance between Ybar and E(Y) is larger than c, goes to 0 and n goes to infinity</p>
28
New cards

the central limit theorem CLT

in an iid sample, Zn is asymptotically normal

→ for any z, P(Zn≤) converges to Φ(z) as n→∞

CLT also requires V(Y) must be strictly positive and finite

29
New cards

estimator

a random variable which is a function of the data sample

30
New cards

estimate

the numerical value of the estimator when computed on a specific sample

31
New cards

bias

when μ^ is an estimator of μ

bias: E(μ^) = μ

an estimator is unbiased if E(μ^) = μ

→ unbiasedness means that is we were to compute μ^ over many different samples, the average would be μ

32
New cards

consistency

μ^ is a consistent estimator is μ^ converges in probability to μ

→ having a consistent estimator in such a large sample means that the estimated value is probably very close to the true value

→ μ^ → p → μ for any c>0, the probability that the distance between μ^ and μ is larger than c goes to 0 as the sample becomes infinitely large

33
New cards

efficiency

μ^ is more efficient than μ^~ if V(μ^) <V(μ^~)

  • the values taken by the more efficient estimator tend to be less dispersed


34
New cards

confidence intervals

if b is a parameter and q a probability (often 95%). the q confidence interval for b CIq(b) contains the true value of b with probability q

→ it’s the set of values b* that we could accept b=b* against b≠b* at the 1-q significance level

in practice with an unbiased and (asymptotically) normal estimator b^ of b:

<p>if <em>b</em> is a parameter and <em>q</em> a probability (often 95%). the <em>q</em> confidence interval for <em>b</em> <em>CI</em>q(b) contains the true value of <em>b </em>with probability <em>q</em></p><p><em>→ </em>it’s the set of values b* that we could accept b=b* against b≠b* at the 1-q significance level</p><p><em>in practice with an unbiased and (asymptotically) normal estimator b^ of b:</em></p>
35
New cards

p-values

the p-value of a test is the largest significance level at which we could conduct the test and still fail to reject Ho

p-value = P Ho[ |T|>|t| ]

  • T is the test statistic, t the realisation in the data sample

  • if the probability is small this value of T is unlikely so we can reject Ho, if large then we cannot


36
New cards

matrices

a matrix M size (r, c) with r rows and c columns

  • each element is denoted mij with i as the row number and j the column number


37
New cards

column and row vectors in matricies

a size (r,1) matrix is a column vector and a row vector is a (1, c) matrix

38
New cards

square matrix

has the same number of rows and columns : r = c

a square matrix M is symmetric if mij = m ji fir (i, j)

39
New cards

commutative

if a.b = b.a which is not true of matrix mutiplication and would only be true if one of the quantities was the identity matrix

40
New cards

transpose of a matrix

denoted as M’ or M^t

such that the kth row equals the kth column of M so if P=M’ elements pij = mji

41
New cards

when will a square matrix be symmetric

if M’ = M

→ original matrix equals transpose if its symmetric

42
New cards

transpose of a product

if the product MP exists then

(MP)’ = P’ M’

43
New cards

product of transposes

let U be a size (n, 1) column vector:

<p>let <em>U</em> be a size (n, 1) column vector: </p>
44
New cards

semi-definite matrices

a square (n, n) matrix M is positive semi-definite if there exists a matrix p with n rows such that:

M=PP’

→ we can think of PP’ as the ‘square’ of matrix P

45
New cards

invertability

a square matrix M (n,n) is invertible is there exists a square matrix, usually denoted M^-1, such that

MM^-1 = M^-1 M = In

  • if M is invertible, M^-1 it its inverse

  • if the inverse of a matrix exists, it is unique

  • if M and P are both invertible matrices, the product MP is invertible:

(MP)^-1 = P^-1 M^-1


46
New cards

linear dependence

a finite number of column vectors of similar size (r, 1) are linearly dependent if one of them can be written as a weighted sum of the other vectors

eg. A + B = 1 or A=2B

47
New cards

matrix rank / full column rank

the maximum number of columns of M that are not linearly dependent

if the rank of a matrix is equal to its column size the matrix is full column rank → if M is full column rank MM’ is invertible

eg. for A (m, n) rank A=n → full column rank

  • a square matrix is invertible if and only if it’s full column rank


48
New cards

random matrix

entries are random variables with a specified joint statistical distribution

the expectation E(X) of matrix X is the (n, K) matrix:

<p>entries are random variables with a specified joint statistical distribution</p><p>the expectation E(X) of matrix X is the (n, K) matrix:</p>
49
New cards

the variance-covariance matrix of a random vector (1, n)

V(Y) = E(YY’) - E(Y) [E(Y)] ‘


→ if α is scalar then V(αY) = α² V(Y)

→ if M is a deterministic/fixed (r,n) matrix (no random elements) then:

V(MY) = M V(Y) M’

<p>V(Y) = E(YY’) - E(Y) [E(Y)] ‘</p><p></p><p>→ if α is scalar then V(αY) = α² V(Y)</p><p>→ if M is a deterministic/fixed (r,n) matrix (<em>no random elements)</em> then:</p><p>V(MY) = M V(Y) M’</p>
50
New cards

SSR (sum of squared residuals) matrix

SSR(b) = (Y-Xb)’(Y-Xb)

<p><strong>SSR(<em>b</em>) = (Y-Xb)’(Y-Xb)</strong></p>
51
New cards

OLS definition (matrix)

the sum of squared residuals SSR(b) is a function of parameter vector b

→ where solving this minimisation problem gives the expression of the OLS estimator

<p>the <em>sum of squared residuals</em> SSR(<em>b</em>) is a function of parameter vector <em>b</em></p><p><em>→ </em>where solving this minimisation problem gives the expression of the OLS estimator</p>
52
New cards

matrix OLS with one regressor (FOCs)

with matrix X and matrix b (1, 2) (containing only b0, b1)

the SSR becomes:

<p>with matrix X and matrix <em>b </em>(1, 2)<em> (containing only b0, b1)</em></p><p>the SSR becomes: </p>
53
New cards

the expression of the OLS estimator (with covariance and variance)

using V^ (x1) = (x1²)bar - (x1bar)² and cov^(y, x1) = (yx1)bar-x1bar.ybar

<p>using V^ (x1) = (x1²)bar - (x1bar)² and cov^(y, x1) = (yx1)bar-x1bar.ybar</p>
54
New cards

matrix OLS with multiple regressors

OLS estimator β^ minimises:


then the FOCs rewrite as:

0=X’ (Y-Xβ^) → where the 0 is actually a (k+1) column vector of zeros


we than then solve for β^ → X’Y-X’Xβ^ = 0

X’Y=X’Xβ^

so, the OLS estimator equals:

β^= (X’X)^-1 X’Y

<p>OLS estimator β^ minimises: </p><p></p><p>then the FOCs rewrite as:</p><p><strong>0=X’ (Y-X</strong>β^)    → where the 0 is actually a (k+1) column vector of zeros</p><p></p><p>we than then solve for β^ → X’Y-X’Xβ^ = 0</p><p>X’Y=X’Xβ^</p><p>so, the OLS estimator equals:</p><p>β^= (X’X)^-1 X’Y</p>
55
New cards

OLS existence and multicollinearity facts

  • OLS estimator exists if the system of FOCs admits a unique solution (invertible)

  • OLS exists if the matrix X’X is invertible

  • the square matrix X’X is invertible if matrix X is full column rank (the (k+1) columns of X must be linearly independent)

  • if these columns are not independent the regressors are multicollinear


56
New cards

predicted outcome in matricies

the predicted outcome is defined as:

yi^ = X’i β^

and the vector of predicted outcomes is

<p>the <em>predicted outcome</em> is defined as:</p><p>y<em>i</em>^ = <em>X</em>’i <span>β^</span></p><p><span>and the <em>vector</em> of predicted outcomes is</span></p>
57
New cards

the vector of predicted residuals

the predicted residual is defined as:

ui^ = yi - yi^ → the error

the vector of predicted residuals is:


<p>the <em>predicted residual </em>is defined as:</p><p>u<em>i</em>^ = y<em>i</em> - y<em>i</em>^   → the error</p><p>the vector of predicted residuals is:</p><p></p>
58
New cards

equation of how the predicted error is uncorrelated with any regressor

knowt flashcard image
59
New cards

TSS (total sum of squares)

TTS is a measure of the dispersion of the outcome of interest

  • TSS/ (n-1) is the sample variance

using yi = yi^ + ui we have:



<p>TTS is a measure of the dispersion of the outcome of interest</p><ul><li><p>TSS/ (n-1)  is the sample variance</p></li></ul><p>using <em>y</em>i = <em>y</em>i^ + <em>u</em>i we have:</p><p></p><p></p>
60
New cards

proof of the equality TSS = ESS + SSR

the last term equals 0 as we know the predicted outcome y^ and residual u^ are uncorrelated, and the mean u^bar of u^ is 0

→

0=cov^ (y^,u^)={sum to n} (yi^-yibar)(ui^-ui^bar)= {sum to n} (yi^-yibar)ui^ = 0

<p>the last term equals 0 as we know the predicted outcome y^ and residual u^ are uncorrelated, and the mean u^bar of u^ is 0 </p><p>→ </p><p>0=cov^ (y^,u^)={sum to n} (yi^-yibar)(ui^-ui^bar)= {sum to n} (yi^-yibar)ui^ = 0</p>
61
New cards

how to determine the fit of OLS (coefficient of determination)

the coefficient of determination is R²

R²=ESS/TSS = 1- SSR/TSS

  • the closer R² is to 1, the better the fit of the OLS regression

eg/ if R² = 0.2 → 80% of the change in Yi is coming from unobserved factors


62
New cards

exogeneity assumption

the exogeneity of the x variables:

E(u| x1, …, xk) = 0

as follows:

E(u)=0, E(u|xk)=0, E(uxk)=0, cov(u, xk)=0

63
New cards

proof of the exogeneity assumption

  1. E(u)=0

E(u)=E[E(u|x1, …, xk)] = E[0] = 0

  1. E(u|x1)=0

E(u|x1)=E[E(u|x1, x2, …, xk)]= E[0]=0

  1. E(ux1)=0

E(ux1)=E[E(ux1|x1)]=E[x1E(u|x1)]=E[x1.0]=0

  1. cov(u, xk)=0

cov(u, xk)=E(uxk)-E(xk)E(u)=0-E(xk).0 =0


64
New cards

moments conditions (2)

due to exogeneity and the solutions to the FOCs:

0=ybar-βo^-β1^x1^

0=(yx1)bar-βo^x1bar-β1^x²1^

we get the moment conditions:

<p>due to exogeneity and the solutions to the FOCs:</p><p><strong>0=ybar-</strong><span><strong>β<em>o</em>^-β<em>1</em>^<em>x</em>1^</strong></span></p><p><strong>0=(y<em>x</em>1)bar-</strong><span><strong>β<em>o</em>^<em>x</em>1bar-β<em>1</em>^<em>x²</em>1^</strong></span></p><p><span>we get the moment conditions:</span></p>
65
New cards
<p>OLS is <strong>unbiased</strong> in finite samples if</p>

OLS is unbiased in finite samples if

assuming an:

  • iid sample

  • no multicolinearity

  • exogeneity : E(ui|Xi)=0 for any i

  • a linear model: yi=X’i β +ui for any i


66
New cards

proof of unbiasedness


67
New cards

variance of the OLS estimator

the average of the square - the square of the average

V(β^| X)=E(β^β^’ | X)-E(β^|X)E(β^|X)’

and plugging in (from the image): we get the general expression for conditional variance of OLS→

V(β^|X)=(X’X)^-1 X’E(UU’| X)X(X’X)^-1

<p>the average of the square - the square of the average</p><p><strong>V(β^| X)=E(β^β^’ | X)-E(β^|X)E(β^|X)’</strong></p><p>and plugging in (<em>from the image): </em>we get the general expression for conditional variance of OLS→</p><p><strong>V(β^|X)=(X’X)^-1 X’E(UU’| X)X(X’X)^-1</strong></p>
68
New cards

difference between homoskedasticity and heteroskedasticy

homoskedasticity: if all its random variables have the same finite variance

heteroskedasticity: if all random variables do not have the same finite variance

  • often occurs when there is a large difference among the size of the observations

→ eg. poor people will always buy cheaper food, but wealthy people with have more variance in eating both expensive and cheaper food


69
New cards

homoskedasticity

V(U|X)=σ²In

→ all error terms are not correlated (conditional on X) and they all have the same conditional variance σ²

combining exogeneity and heteroskedasticity we have:

E(UU’|X)=σ²In

plugging this into the OLS variance formula we get:

V(β^|X)=σ²(X’X)^-1

70
New cards

consistency of OLS → and what this means:

under exogeneity, OLS is a consistent estimator of β:

if β^ → p → β as β^ converges in probability to β

the probability there is a gap between β^ and β → 0 as n→ထ

  • the estimator of the constant converges to the model constant

βo^ →p→βo

  • for any regressor xk, the estimated coefficient converges to the causal effect:

βk^ → p →βk


71
New cards

asymptotic normality

an estimator that is asymptotically normal will have an approximately normal distribution as the sample size gets infinitely large

→ the distribution of an estimator becomes more normal as the n tends to infiinity

72
New cards

proof of asymptotic normality

from the proof of consistency we know (image):

  • where 1/n X’X converges to a matrix of expectations (not asymptotically normal)

  • and 1/n X’U is a vector of sample means

→ by CLT, each element of 1/n X’U is asymptotically normal

so β^ is a linear transformation of an asymptotically normal variable and is thus asymptotically normal

→ as a constant times a normal is still normal


<p>from the proof of consistency we know (<em>image):</em></p><ul><li><p>where <strong>1/n X’X </strong>converges to a matrix of expectations (<em>not asymptotically normal)</em></p></li><li><p>and <strong>1/n X’U </strong>is a vector of sample means </p></li></ul><p>→ by CLT, each element of <strong>1/n X’U is asymptotically normal</strong></p><p><em>so </em>β^ is a linear transformation of an asymptotically normal variable <em>and is thus</em> asymptotically normal</p><p>→ as a constant times a normal is still normal</p><p></p>
73
New cards

when to use different tests (T or F)?

we can conduct tests on one parameter βk at a time→ T test

we can conduct tests on combinations of parameters → F test

74
New cards

T test example

consider a parameter in β: βk and a constant c →

test Ho: βk = c , Ha: βk

using the following test statistic (image):

  • the test statistic T and its asymptotic distribution are the same one sided vs. two sided

  • in a large sample , T is approximately distributed as a standard normal


<p>consider a parameter in β<span><em>:  </em></span>β<span><em>k </em>and a constant c →</span></p><p><span>test Ho: </span>β<span><em>k = c , </em>H<em>a</em>: </span>β<span><em>k</em></span></p><p><span><em>using the following test statistic </em>(image):</span></p><ul><li><p>the test statistic <em>T</em> and its asymptotic distribution are the same one sided vs. two sided</p></li><li><p>in a large sample , T is approximately distributed as a standard normal</p></li></ul><p></p>
75
New cards

F test example

testing on multiple parameters eg. Qβ=C where Q is a (q,k+1) matrix and C a (q, 1)

{ eg. βmale = 2βage}

test Ho: Qβ=C, Ha: Qβ≠C

the test statistic is:


<p>testing on multiple parameters eg. <em>Qβ=C where Q is a (q,k+1) matrix and C a (q, 1)</em></p><p><em>{ eg. </em><span><em>βmale = 2βage}</em></span></p><p><span>test Ho: Qβ=C, H<em>a: </em></span>Qβ≠C</p><p><em>the test statistic is: </em></p><p></p>
76
New cards

internal validity of OLS (2 ways this can be violated0

  • the estimator of causal effects of interest may be biased or not consistent (exogeneity doesn’t hold)

  • the estimated standard errors may be inconsistent (sample not iid/ heteroskedasticity)


77
New cards

omitted variable bias

consider linear model y=bo + b1x1 + u

assume there is an unobserved variable x2 not included in y that is:

  • correlated with x1 and y

thus, it lies in the error term

u=b2x2 + ϵ where E(ϵ| x1)=0

OLS is thus biased and inconsistent for the causal paramter b1

  • if x2 is positively correlated with y → bias is positive, and y is overestimated, and vice versa


78
New cards

simultaneity/ reverse causality

Y could also have a causal effect on X

linear model: y=bo + b1x1 + u →

reverse causality arises from: x1 = ao + a1y + v (where cov(v, u)=0)

as x1 and u are correlated exogeneity is violated (image)

<p>Y could also have a causal effect on X </p><p><em>linear model: </em><strong>y=bo + b1x1 + u </strong> →</p><p><em>reverse causality arises from: </em><strong>x1 = ao + a1y + v</strong>  (<em>where </em> cov(v, u)=0)</p><p>as x1 and u are <strong>correlated</strong> <em>exogeneity is violated </em>(image)</p>
79
New cards

selection bias

arises when some data are missing as a result of a selection process that depends on Y, even after controlling for the regressors

→ Y= wage, x1=women yet not all women participate in labour market, so sub-sample of those who do may have different characteristics to the larger sample of the population , so the OLS estimate on the sample will be biased

80
New cards

can we regress OLS Y on a constant, m, x and x² ?

we can run such a regression as the relationship between age and age² is not linear so there is no multicollinearity issue, as long as age(or gender) is not constant in the sample

  • OLS fails if there is perfect linear dependence eg. if a²= c +da , as it would make X’X singular

  • eg. if a is constant at 30 (everyone in sample is 30) then age is perfectly collinear with the intercept. so OLS would fail because the regressors are no longer linearly independent


81
New cards

what does it mean if the OLS estimator is BLUE

this is the Gauss Markov theorem

BLUE stands for best (most efficient, lowest variance), linear (OLS is a linear transformation of the outcome vector), unbiased (the expectation of the OLS estimator is the vector of causal parameters)

  • under exogeneity and homoskedasticity, OLS has the lowest variance in the class of unbiased linear estimators


82
New cards

two properties for an instrumental variable to be valid

  1. Instrumental relevance

cov(z, x) ≠ 0

z must be correlated to the endogenous explanatory variable x

  1. Instrumental exogeneity

cov(z, u) =0

and must be exogenous in the linear model


83
New cards

setting up the instrument z with x

x = ao + a1z + ϵ

and predict

x^ = ao^+a1^z

84
New cards

two-stage least squares estimator 2SLS

if we have a valid instrument x we can use 2SLS , obtained by two consecutive OLS regressions

85
New cards

formula for t test in ols

T=(β-0) / SE

where SE = root var

86
New cards

F-statistic measures :

the relative increase in SSR per additional predictor, adjusted for the degrees of freedom

  • useful when testing the collective significance of related predictors x1, x2,… or dummy variables


87
New cards

null and alternative hypothesis in F-test

null hypothesis, Ho, states that the group of predictors tested jointly have no significant effect

eg. β1=β2=…βk=0

alternative hypothesis, H1, states that at least one of the predictors has a significant effect on the dependent variable

eg. at least one βi ≠ 0

88
New cards

the test-statistic calculation in F-test

SSRr and SSRur are the sum of squared residuals for the restricted/ unrestriced models

dfr and dfur are their respective degrees of freedom

  • the critical value is determined by the chosen significance level (α) and the degrees of freedom

  • of the F-statistic exceeds the critical value, or p<α, the null hypothesis is rejected → indicating that the group of predictors jointly have a significant effect on the dependent variable


<p>SSR<em>r </em>and SSR<em>ur </em> are the sum of squared residuals for the restricted/ unrestriced models</p><p>df<em>r </em>and df<em>ur </em>are their respective degrees of freedom</p><ul><li><p>the critical value is determined by the chosen significance level (<span>α) and the degrees of freedom</span></p></li><li><p><span>of the F-statistic exceeds the critical value, or p&lt;α, the null hypothesis is rejected → indicating that the group of predictors jointly have a significant effect on the dependent variable</span></p></li></ul><p></p>
89
New cards

if u^ is the predicted residual, provde V(u^ | m,p) = E(u^²|m,p)

rewrite u^ as the fitted linear model of

then replace w with the linear model

using the exogeneity assumption that E(u| m,p)=0

<p>rewrite u^ as the  fitted linear model of </p><p>then replace w with the linear model</p><p>using the exogeneity assumption that E(u| m,p)=0 </p>
90
New cards

show how testing for two slope parameters in this regression being jointly equal can inform on heteroskedasticity

  • an F-test will tell us if the coefficients of m and p are jointly equal to 0

  • if we reject Ho, either m or p has some impact on the squared precited residual [u^²=βo+β1x1+β2


91
New cards

first stage of the 2SLS

isolate the part of x that is uncorrelated with u. regress x on x using OLS

x = ao + a1z + ϵ

as x is uncorrelated with u, so its ao and a1z

yielding two-parameter estimates ao^, a1^, so we check instrument relevence : Ho ; a1 = 0 , H1: a1 ≠ 0

compute the predicted values of x (which captures the exogenous part of x), denoted by x^ as: x^=ao^ +a1^z

the only source of variation in x^ comes from z and is thus exogenous:

cov(x^,u)=cov(ao^+a1z, u)=a1^cov(z,u)=0

92
New cards

second stage of the 2SLS estimator

regress y on the predicted regressor x^ by OLS

  • because x^ is uncorrelated with u, β1 can be estimated by OLS

  • we look at how y responds to exogenous changes in x^

→ this can be interpreted as a causal response

  • the resulting estimator is the two-stage least squared estimator


<p>regress y on the predicted regressor x^ by OLS</p><ul><li><p>because x^ is uncorrelated with u, β1 can be estimated by OLS</p></li><li><p>we look at how y responds to <em>exogenous</em> changes in x^ </p></li></ul><p>→ this can be interpreted as a causal response</p><ul><li><p>the resulting estimator is the <em>two-stage least squared estimator</em></p></li></ul><p></p>
93
New cards

consistency of the 2SLS estimator

y=βo +β1x+u

we can rewrite β1 in terms of population moments:

cov(y,z)=cov(βo+β1x+u, z)=β1cov(x, z)

where cov(u,z)=0 due to instrument exogeneity (yet if the instrument is not relevant, then β1 is not defined)

rearranging we can express β1 as

by the law of large numbers: the 2SLS estimator is consistent for β1 if z is a valid instrument for x:

B^(2SLS) → p → β1 (if either of the two assumptions needed for a valid instrument fails, the 2SLS estimator will not be consistent)

<p>y=βo +β1x+u</p><p>we can rewrite β1 in terms of population moments:</p><p><strong>cov(y,z)=cov(βo+β1x+u, z)=β1cov(x, z)</strong></p><p>where cov(u,z)=0 due to <em>instrument exogeneity </em>(yet if the instrument is <strong>not </strong>relevant, then β1  is <strong>not</strong> defined)</p><p>rearranging we can express β1 as</p><p>by the law of large numbers: the 2SLS estimator is consistent for β1 if z is a valid instrument for x:</p><p><strong>B^(2SLS) → p → β1 (</strong>if either of the two assumptions needed for a valid instrument fails, the 2SLS estimator will not be consistent)</p>
94
New cards

the 2SLS is not unbiased when:

  • x and u are correlated → meaning, when instrumental variable estimation is actually needed, the 2SLS estimator is essentially never unbiased


95
New cards

properties of the 2SLS estimator

if we have a large iid sample and a valid instrument, the 2SLS estimator β^(2SLS) has an approximately normal distribution

  • with mean β1, following from the consistency result

  • with a variance that has a (complicated) closed-form expression → statistical software will estimate the standard errors


96
New cards

alternative derivation of 2SLS estimator

start with y=βo+β1x+u , where E(u|x)≠0

use reduced form to relate y to z:

y=γo+γ1z+μ

the first-stage related x to z

x=πo+π1z+ε → where μ and ε are error terms, as we assume x is exogenous and uncorrelated with both

solve the first stage for x and substitute it in the reduced form

—> the effect on y of an exogenous change in x is β1=γ1 / π1 units

<p>start with <em>y=</em><span><em>β</em></span>o+<span><em>β</em></span>1x+u   , <em>where E(u|x)</em><span>≠0</span></p><p><span>use reduced form to relate y to z:</span></p><p><span>y=γo+γ1z+μ</span></p><p><span>the <em>first-stage</em> related x to z</span></p><p><span>x=</span>πo+π1z+ε  → where <span>μ </span>and ε are error terms, as we assume x is exogenous and uncorrelated with both</p><p>solve the first stage for x and substitute it in the reduced form</p><p>—&gt; the effect on <em>y</em> of an exogenous change in <em>x</em> is <span><em>β</em></span>1=<span><em>γ</em></span>1 / <span><em>π</em></span>1 units</p>
97
New cards

identification of the coefficients of IV regression

whether the coefficients are identified depends on the no. instruments (m) and the no. endogenous regressors (k)

so the coefficients β1,..,βk are said to be:

  • exactly identified: if m=k

there are just enough instruments to estimate β1,..,βk

  • overidentified: if m>k

there are more than enough instruments to estimate β1,…,βk

we can test whether the instruments are valid

  • underidentified: if m<k

there are not enough instruments to estimate β1,..,βk, if so, you need to get more instruments


98
New cards
<p>the 2SLS estimator in matrix form </p>

the 2SLS estimator in matrix form

we exploit the fact that P is independent and symmpetric so P’=P and PP=P

and X^=XP

99
New cards

showing that the 2SLS estimator is not unbiased in matrix form

Y=Xβ + U and plug it into the 2SLS estimator

  • to prove unbiasedness, we would need to show the last term is 0 in expectation

  • we would use expectations conditional on X, as under exogeneity E(U|X)=0

  • under instrument exogeneity, E(U|Z)=0 yet this is insufficient as the last term also involves the random variables X


<p><strong>Y=X</strong><span><strong>β</strong></span><strong> + U  </strong>and plug it into the 2SLS estimator </p><ul><li><p>to prove unbiasedness, we would need to show the <strong>last term is 0 in expectation</strong></p></li><li><p>we would use expectations conditional on X, as under exogeneity E(U|X)=0</p></li><li><p>under instrument exogeneity, E(U|Z)=0 yet this is insufficient as the <strong>last term</strong> also involves the random variables X</p></li></ul><p></p>
100
New cards

our instrument is “strong” if

the F-test for it is larger than 10

  • instruments are weak if they explain very little of the variation in x, beyond the one explained by the w’s (included exogenous variables in z) → if all the π1,..,πm are either 0 or nearly 0

  • too weak correlation between x and z can cause the 2SLS estimator to have a large asymptotic bias even if z and u are only moderately correlated

2SLS will perform worse than OLS (will have a larger asymptotic bias) if:


<p>the F-test for it is larger than 10</p><ul><li><p>instruments are weak if they explain very little of the variation in x, beyond the one explained by the w’s (included exogenous variables in z) → <em>if all the </em><span><em>π</em></span>1,..,<span><em>π</em></span>m <em>are either 0 or nearly 0</em></p></li><li><p>too weak correlation between x and z can cause the 2SLS estimator to have a <strong>large asymptotic bias</strong> even if z and u are only moderately correlated</p></li></ul><p>2SLS will perform worse than OLS (<em>will have a larger asymptotic bias) </em>if:</p><p></p>