Multiple Regression Analysis

y = 0 + 1x1 + 2x2 + . . . kxk + u

1. Estimation

Economics 20 - Prof. Anderson 1

Parallels with Simple Regression
0 is still the intercept
1 to k all called slope parameters
u is still the error term (or disturbance)
Still need to make a zero conditional mean
assumption, so now assume that
E(u|x1,x2, …,xk) = 0
Still minimizing the sum of squared
residuals, so have k+1 first order conditions
Economics 20 - Prof. Anderson 2
Interpreting Multiple Regression

yˆ  ˆ0  ˆ1 x1  ˆ 2 x2  ...  ˆ k xk , so

yˆ  ˆ x  ˆ x  ...  ˆ x ,
1 1 2 2 k k

so holding x2 ,..., xk fixed implies that

yˆ  ˆ x , that is each  has
1 1

a ceteris paribus interpretation

Economics 20 - Prof. Anderson 3

A “Partialling Out” Interpretation
Consider the case where k  2, i.e.
yˆ  ˆ  ˆ x  ˆ x , then
0 1 1 2 2

ˆ1    rˆi1 yi   rˆ 2
i1 , where rˆi1 are
the residuals from the estimated
regression xˆ1  ˆ0  ˆ2 xˆ 2
Economics 20 - Prof. Anderson 4
“Partialling Out” continued
Previous equation implies that regressing y
on x1 and x2 gives same effect of x1 as
regressing y on residuals from a regression of
x1 on x2
This means only the part of xi1 that is
uncorrelated with xi2 are being related to yi so
we’re estimating the effect of x1 on y after x2
has been “partialled out”
Economics 20 - Prof. Anderson 5
Simple vs Multiple Reg Estimate

~ ~ ~
Compare the simple regression y   0  1 x1
with the multiple regression yˆ  ˆ0  ˆ1 x1  ˆ 2 x2
~
Generally, 1  ˆ1 unless :
ˆ  0 (i.e. no partial effect of x ) OR
2 2

x1 and x2 are uncorrelated in the sample

Economics 20 - Prof. Anderson 6

Goodness-of-Fit
We can think of each observation as being made
up of an explained part, and an unexplained part,
yi  yˆ i  uˆi We then define the following :
  y  y  is the total sum of squares (SST)
2
i

  yˆ  y  is the explained sum of squares (SSE)

2
i

 uˆ is the residual sum of squares (SSR)

2
i

Then SST  SSE  SSR

Economics 20 - Prof. Anderson 7
Goodness-of-Fit (continued)
How do we think about how well our
sample regression line fits our sample data?

Can compute the fraction of the total sum

of squares (SST) that is explained by the
model, call this the R-squared of regression

R2 = SSE/SST = 1 – SSR/SST

Economics 20 - Prof. Anderson 8

Goodness-of-Fit (continued)
We can also think of R 2 as being equal to
the squared correlation coefficient between
the actual yi and the values yˆ i
   y  y   yˆ  yˆ  
2

  y  y    yˆ  yˆ  
i i
R 2
 2 2
i i

Economics 20 - Prof. Anderson 9

More about R-squared
R2 can never decrease when another
independent variable is added to a
regression, and usually will increase

Because R2 will usually increase with the

number of independent variables, it is not a
good way to compare models

Economics 20 - Prof. Anderson 10

Assumptions for Unbiasedness
Population model is linear in parameters: y = 0 +
1x1 + 2x2 +…+ kxk + u
We can use a random sample of size n, {(xi1, xi2,…,
xik, yi): i=1, 2, …, n}, from the population model, so
that the sample model is yi = 0 + 1xi1 + 2xi2 +…+
kxik + ui
E(u|x1, x2,… xk) = 0, implying that all of the
explanatory variables are exogenous
None of the x’s is constant, and there are no exact
linear relationships among them

Economics 20 - Prof. Anderson 11

Too Many or Too Few Variables
What happens if we include variables in
our specification that don’t belong?
There is no effect on our parameter
estimate, and OLS remains unbiased

What if we exclude a variable from our

specification that does belong?
OLS will usually be biased

Economics 20 - Prof. Anderson 12

Omitted Variable Bias
Suppose the true model is given as
y   0  1 x1   2 x2  u , but we
~ ~ ~
estimate y     x  u , then
0 1 1

~
1 
 xi1  x1  yi
 x i1  x1 
2

Economics 20 - Prof. Anderson 13

Omitted Variable Bias (cont)
Recall the true model, so that
yi   0  1 xi1   2 xi 2  ui , so the
numerator becomes
  x  x  
i1 1 0  1 xi1   2 xi 2  ui  
  x  x    2   xi1  x1  xi 2    xi1  x1 ui
2
1 i1 1

Economics 20 - Prof. Anderson 14

Omitted Variable Bias (cont)

~
  1   2  x  x x   x
i1 1 i2 i1  x1  ui
  x  x     x  x1  
2 2
i1 1 i1

since E(ui )  0, taking expectations we have

 
~
E  1  1   2
 x  x x i1 1 i2

  x  x  
2
i1 1

Economics 20 - Prof. Anderson 15

Omitted Variable Bias (cont)

Consider the regression of x2 on x1

~ ~ ~ ~
x2   0   1 x1 then  1 
 x  x x
i1 1 i2

  x  x  
2
i1 1

 
~ ~
so E 1  1   2 1

Economics 20 - Prof. Anderson 16

Summary of Direction of Bias

Corr(x1, x2) > 0 Corr(x1, x2) < 0

2 > 0 Positive bias Negative bias

2 < 0 Negative bias Positive bias

Economics 20 - Prof. Anderson 17

Omitted Variable Bias Summary
Two cases where bias is equal to zero
 2 = 0, that is x2 doesn’t really belong in model
 x1 and x2 are uncorrelated in the sample

If correlation between x2 , x1 and x2 , y is

the same direction, bias will be positive
If correlation between x2 , x1 and x2 , y is
the opposite direction, bias will be negative
Economics 20 - Prof. Anderson 18
The More General Case
Technically, can only sign the bias for the
more general case if all of the included x’s
are uncorrelated

Typically, then, we work through the bias

assuming the x’s are uncorrelated, as a
useful guide even if this assumption is not
strictly true

Economics 20 - Prof. Anderson 19

Variance of the OLS Estimators
Now we know that the sampling distribution
of our estimate is centered around the true
parameter
Want to think about how spread out this
distribution is
Much easier to think about this variance under
an additional assumption, so
Assume Var(u|x1, x2,…, xk) = 2
(Homoskedasticity)
Economics 20 - Prof. Anderson 20
Variance of OLS (cont)
Let x stand for (x1, x2,…xk)
Assuming that Var(u|x) = 2 also implies
that Var(y| x) = 2

The 4 assumptions for unbiasedness, plus

this homoskedasticity assumption are
known as the Gauss-Markov assumptions

Economics 20 - Prof. Anderson 21

Variance of OLS (cont)
Given the Gauss - Markov Assumptions

   2
Var ˆ j  , where
SST j 1  R j 
2

SST j    xij  x j  and R is the R

2 2 2
j

from regressing x j on all other x' s

Economics 20 - Prof. Anderson 22
Components of OLS Variances
The error variance: a larger 2 implies a
larger variance for the OLS estimators
The total sample variation: a larger SST j
implies a smaller variance for the estimators
Linear relationships among the
independent variables: a larger Rj2 implies a
larger variance for the estimators

Economics 20 - Prof. Anderson 23

Misspecified Models

Consider again the misspecified model


 
2
~ ~ ~ ~
y   0  1 x1 , so that Var 1 
SST1
 
~
 
Thus, Var   Var ˆ unless x and
1 1 1

x2 are uncorrelated, then they' re the same

Economics 20 - Prof. Anderson 24

Misspecified Models (cont)
While the variance of the estimator is
smaller for the misspecified model, unless
2 = 0 the misspecified model is biased

As the sample size grows, the variance of

each estimator shrinks to zero, making the
variance difference less important

Economics 20 - Prof. Anderson 25

Estimating the Error Variance
We don’t know what the error variance, 2,
is, because we don’t observe the errors, ui

What we observe are the residuals, ûi

We can use the residuals to form an

estimate of the error variance

Economics 20 - Prof. Anderson 26

Error Variance Estimate (cont)
ˆ    uˆ
2
  n  k  1  SSR df
2
i

thus, se ˆ   ˆ  SST 1  R  

j j
2 12
j

df = n – (k + 1), or df = n – k – 1
df (i.e. degrees of freedom) is the (number
of observations) – (number of estimated
parameters)
Economics 20 - Prof. Anderson 27
The Gauss-Markov Theorem
Given our 5 Gauss-Markov Assumptions it
can be shown that OLS is “BLUE”
Best
Linear
Unbiased
Estimator
Thus, if the assumptions hold, use OLS

Economics 20 - Prof. Anderson 28