DSOM 309 CHAPTER 16 EXAM QUESTIONS AND ANSWERS
WITH COMPLETE SOLUTIONS VERIFIED LATEST UPDATE
these can be used to compare simple linear regression that have
same response variables
R² and se
if
y=β₀+β₁ln(x)+e
obtain an estimate for β₁ of -500
what does the value imply
if x increases by 1% then y is expected to decrease by 5 units
B₁/100 (-500/100=/5)
the linearity assumption implys that if x goes up by one unit
we expect y to change by β₁, irrespective of value of x
in a cubic regression model how many times does the sign of the
slope change
two
the coefficient of determination R² cannot be used to compare the
linear and quadratic models bc
the quadratic has one parameter more to estimate
ln(y)=β₀+β₁ln(x)+e
this is a
log-log regression model
, yhat=b₀+b₁x
the marginal effect of x on yhat is
b₁
which semi-log model transforms only the explanatory variable
logarithmic model
y=β₀+β₁ln(x)+e
which semi-log model transforms only the response variable
exponential model
simple calculation of the coefficient of determination [R²]
is to square the sample
correlation coefficient between y and yhat
in model
ln(y)=β₀+β₁x+e
the term β₁×100 is the approximate
percentage change in E(y) when x increases by 1 unit
in equation
yhat=b₀+b₁x+b₂x²
the marginal effect of x on yhat is
b₁+2b₂x
suppose you estimate
ln(y)=β₀+β₁ln(x)+e
and obtain an estimate of β₁ of -2.4
what does this value imply
WITH COMPLETE SOLUTIONS VERIFIED LATEST UPDATE
these can be used to compare simple linear regression that have
same response variables
R² and se
if
y=β₀+β₁ln(x)+e
obtain an estimate for β₁ of -500
what does the value imply
if x increases by 1% then y is expected to decrease by 5 units
B₁/100 (-500/100=/5)
the linearity assumption implys that if x goes up by one unit
we expect y to change by β₁, irrespective of value of x
in a cubic regression model how many times does the sign of the
slope change
two
the coefficient of determination R² cannot be used to compare the
linear and quadratic models bc
the quadratic has one parameter more to estimate
ln(y)=β₀+β₁ln(x)+e
this is a
log-log regression model
, yhat=b₀+b₁x
the marginal effect of x on yhat is
b₁
which semi-log model transforms only the explanatory variable
logarithmic model
y=β₀+β₁ln(x)+e
which semi-log model transforms only the response variable
exponential model
simple calculation of the coefficient of determination [R²]
is to square the sample
correlation coefficient between y and yhat
in model
ln(y)=β₀+β₁x+e
the term β₁×100 is the approximate
percentage change in E(y) when x increases by 1 unit
in equation
yhat=b₀+b₁x+b₂x²
the marginal effect of x on yhat is
b₁+2b₂x
suppose you estimate
ln(y)=β₀+β₁ln(x)+e
and obtain an estimate of β₁ of -2.4
what does this value imply
