Lecture 7
Lecture 7
Lecture 7
Note: Statistical Inference is a vast literature (over 200 year old) and is
the subject of numerous textbooks. In this course, we will only focus on a
small part of it.
Note: Statistical Inference is a vast literature (over 200 year old) and is
the subject of numerous textbooks. In this course, we will only focus on a
small part of it.
y = β0 + β1 x + u (1)
y = β0 + β1 x + u (1)
The variable u in eq. (1) is called the error term or disturbance term. It
represents factors other than x that affect y i.e. the unobserved factors.
The variable u in eq. (1) is called the error term or disturbance term. It
represents factors other than x that affect y i.e. the unobserved factors.
∆y = β1 ∆x
Thus the change in y is simply β1 multiplied by change in x. β1 is the
slope parameter. This slope parameter is of primary importance in
econometrics. β0 is the intercept parameter (also called constant term)
although not central to econometrics analysis it is very useful in some
cases.
E (u) = 0 (2)
This assumption simply tells us about the distribution of unobserved
factors in the population, nothing about the relation between x and u.
E (u) = 0 (2)
This assumption simply tells us about the distribution of unobserved
factors in the population, nothing about the relation between x and u.
E (u) = 0 (2)
This assumption simply tells us about the distribution of unobserved
factors in the population, nothing about the relation between x and u.
We assume that the average value of u does not depend on the value of x.
We assume that the average value of u does not depend on the value of x.
Note: Eq (5) tells us how the average value of y changes with x. It does
not say that y equals β0 + β1 x for all units in the population.
Note: Eq (5) tells us how the average value of y changes with x. It does
not say that y equals β0 + β1 x for all units in the population.
yi = β0 + β1 xi + ui (6)
for each i. Here ui is the error term for observation i because it contains
factors other the xi affecting yi .
Therefore u has 0 expected value and that the covariance between x and u
is zero.
Therefore u has 0 expected value and that the covariance between x and u
is zero.
E (u) = 0 (7)
Therefore u has 0 expected value and that the covariance between x and u
is zero.
E (u) = 0 (7)
Therefore u has 0 expected value and that the covariance between x and u
is zero.
E (u) = 0 (7)
Given a sample of data, we choose estimates βˆ0 and βˆ1 to solve the
sample counterparts of eq (9) and eq(10).
n
1X
(yi − βˆ0 − βˆ1 xi ) = 0 (11)
n
i=1
Given a sample of data, we choose estimates βˆ0 and βˆ1 to solve the
sample counterparts of eq (9) and eq(10).
n
1X
(yi − βˆ0 − βˆ1 xi ) = 0 (11)
n
i=1
n
1X
xi (yi − βˆ0 − βˆ1 xi ) = 0 (12)
n
i=1
These two equations (11) and (12) can be solved to get βˆ0 and βˆ1 .
Pn
(x − x̄)(yi − ȳ )
ˆ
β1 = i=1 Pn i 2
(13)
i=1 (xi − x̄)
These two equations (11) and (12) can be solved to get βˆ0 and βˆ1 .
Pn
(x − x̄)(yi − ȳ )
ˆ
β1 = i=1 Pn i 2
(13)
i=1 (xi − x̄)