ISYE 6414 Final Exam Review Exam
Questions and Answers with Verified
Solutions | Latest Updated 2026
Least Square Elimination (LSE) False - it is applicable but does not use
cannot be applied to GLM models. data
distribution information fully.
In multiple linear regression with True - the least squares estimates are
idd BLUE (Best
and equal variance, the least Linear Unbiased Estimates) in multiple
squares linear
estimation of regression regression.
coefficients
are always unbiased.
Maximum Likelihood Estimation is False - In SLR and MLR, the SLE and
not applicable for simple linear MLE are the
regression and multiple linear same with normal idd data.
regression.
The backward elimination requires False - Type I error
a
pre-set probability of type II error
The first degree of freedom in the True
F
distribution for any of the three
procedures in stepwise is always
equal to one.
,MLE is used for the GLMs for True
handling complicated link function
modeling in the X-Y relationship.
In the GLMs the link function False - It can be linear, non linear, or
cannot parametric
be a non linear regression.
When the p-value of the slope False - When P value is small, the model
estimate in the SLR is small the r- fits
squared becomes smaller too. become more significant and R squared
become
larger.
In GLMs the main reason one False - The potential constraint in the
does parameters of
not use LSE to estimate model GLMs is handled by the link function.
parameters is the potential
constrained in the parameters.
The R-squared and adjusted R- TRUE - The underlying assumption of
squared are not appropriate model R-squared
comparisons for non linear calculations is that you are fitting a linear
regression but are for linear model.
regression models.
The decision in using ANOVA True
table
for testing whether a model is
significant depends on the normal
distribution of the response
variable
, When the data may not be True
normally
distributed, AIC is more
appropriate
for variable selection than adjusted
R-squared
The slope of a linear regression False - the correlation coefficient is the r
equation is an example of a value. Will
correlation coefficient. have the same + or - sign as the slope.
In multiple linear regression, as False - r squared measures how much
the variability is
value of R-squared increases, the explained by the model, NOT how strong
relationship the
between predictors becomes predictors are.
stronger
When dealing with a multiple linear False - the adjusted rsquared value take
regression model, an adjusted R- the
squared can number and types of predictors into
be greater than the corresponding account. It is
unadjusted R-Squared value. lower than the r squared value.
In a multiple regression problem, a True
quantitative input variable x is
replaced by x −
mean(x). The R-squared for the
fitted
model will be the same
Questions and Answers with Verified
Solutions | Latest Updated 2026
Least Square Elimination (LSE) False - it is applicable but does not use
cannot be applied to GLM models. data
distribution information fully.
In multiple linear regression with True - the least squares estimates are
idd BLUE (Best
and equal variance, the least Linear Unbiased Estimates) in multiple
squares linear
estimation of regression regression.
coefficients
are always unbiased.
Maximum Likelihood Estimation is False - In SLR and MLR, the SLE and
not applicable for simple linear MLE are the
regression and multiple linear same with normal idd data.
regression.
The backward elimination requires False - Type I error
a
pre-set probability of type II error
The first degree of freedom in the True
F
distribution for any of the three
procedures in stepwise is always
equal to one.
,MLE is used for the GLMs for True
handling complicated link function
modeling in the X-Y relationship.
In the GLMs the link function False - It can be linear, non linear, or
cannot parametric
be a non linear regression.
When the p-value of the slope False - When P value is small, the model
estimate in the SLR is small the r- fits
squared becomes smaller too. become more significant and R squared
become
larger.
In GLMs the main reason one False - The potential constraint in the
does parameters of
not use LSE to estimate model GLMs is handled by the link function.
parameters is the potential
constrained in the parameters.
The R-squared and adjusted R- TRUE - The underlying assumption of
squared are not appropriate model R-squared
comparisons for non linear calculations is that you are fitting a linear
regression but are for linear model.
regression models.
The decision in using ANOVA True
table
for testing whether a model is
significant depends on the normal
distribution of the response
variable
, When the data may not be True
normally
distributed, AIC is more
appropriate
for variable selection than adjusted
R-squared
The slope of a linear regression False - the correlation coefficient is the r
equation is an example of a value. Will
correlation coefficient. have the same + or - sign as the slope.
In multiple linear regression, as False - r squared measures how much
the variability is
value of R-squared increases, the explained by the model, NOT how strong
relationship the
between predictors becomes predictors are.
stronger
When dealing with a multiple linear False - the adjusted rsquared value take
regression model, an adjusted R- the
squared can number and types of predictors into
be greater than the corresponding account. It is
unadjusted R-Squared value. lower than the r squared value.
In a multiple regression problem, a True
quantitative input variable x is
replaced by x −
mean(x). The R-squared for the
fitted
model will be the same
