ISYE 6402 MIDTERM EXAM FULL LENGTH TEST BANK
2025/2026 | ACCURATE CURRENTLY TESTING EXAM
VERSIONS WITH ACTUAL QUESTIONS AND ANSWERS WITH
A RATIONALES AND A STUDY GUIDE| LATEST UPDATE
1. We can assess the constant variance assumption in linear regression by plotting the residuals
vs. fitted values.
A. False
B. True
Rationale: The residual-vs-fitted plot is the standard diagnostic for checking homoscedasticity; a
funnel pattern indicates violation.
2. If one confidence interval in the pairwise comparison in ANOVA includes zero, we conclude
that the two corresponding means are plausibly equal.
A. False
B. True
Rationale: Zero within the CI means no significant difference between those means at the given
confidence level.
3. The assumption of normality is not required in linear regression to make inference on the
regression coefficients.
A. True
B. False
,Rationale: Normality of errors is needed for t- and F-based inference; otherwise, p-values and
CIs are unreliable.
4. We cannot estimate a multiple linear regression model if the predicting variables are linearly
independent.
A. True
B. False
Rationale: Independence is desired; the problem arises when predictors are linearly dependent
(perfect multicollinearity).
5. If a predicting variable is categorical with 5 categories in a regression model without intercept,
we include 5 dummy variables.
A. False
B. True
Rationale: Without an intercept, each category gets its own dummy indicator.
6. If the normality assumption does not hold for regression, we may use a transformation on the
response variable.
A. False
B. True
Rationale: Response transformations (e.g., log, Box-Cox) can stabilize variance and
approximate normality.
7. The prediction of the response variable has higher uncertainty than the estimation of the mean
response.
, A. False
B. True
Rationale: Predictions for new observations include both estimation error and inherent random
error, increasing uncertainty.
8. Statistical inference for linear regression under normality relies on large sample size.
A. True
B. False
Rationale: With normal errors, inference is exact regardless of sample size; large samples matter
mainly for CLT approximations.
9. A nonlinear relationship between response and predictor cannot be modeled using regression.
A. True
B. False
Rationale: Polynomial and transformed predictors allow modeling of nonlinear patterns within a
linear framework.
10. Assumption of normality in linear regression is required for confidence intervals, prediction
intervals, and hypothesis testing.
A. False
B. True
Rationale: These inferential tools rely on normality of residuals for valid t and F distributions.
11. If the confidence interval for a regression coefficient contains zero, we interpret that the
coefficient is plausibly equal to zero.
2025/2026 | ACCURATE CURRENTLY TESTING EXAM
VERSIONS WITH ACTUAL QUESTIONS AND ANSWERS WITH
A RATIONALES AND A STUDY GUIDE| LATEST UPDATE
1. We can assess the constant variance assumption in linear regression by plotting the residuals
vs. fitted values.
A. False
B. True
Rationale: The residual-vs-fitted plot is the standard diagnostic for checking homoscedasticity; a
funnel pattern indicates violation.
2. If one confidence interval in the pairwise comparison in ANOVA includes zero, we conclude
that the two corresponding means are plausibly equal.
A. False
B. True
Rationale: Zero within the CI means no significant difference between those means at the given
confidence level.
3. The assumption of normality is not required in linear regression to make inference on the
regression coefficients.
A. True
B. False
,Rationale: Normality of errors is needed for t- and F-based inference; otherwise, p-values and
CIs are unreliable.
4. We cannot estimate a multiple linear regression model if the predicting variables are linearly
independent.
A. True
B. False
Rationale: Independence is desired; the problem arises when predictors are linearly dependent
(perfect multicollinearity).
5. If a predicting variable is categorical with 5 categories in a regression model without intercept,
we include 5 dummy variables.
A. False
B. True
Rationale: Without an intercept, each category gets its own dummy indicator.
6. If the normality assumption does not hold for regression, we may use a transformation on the
response variable.
A. False
B. True
Rationale: Response transformations (e.g., log, Box-Cox) can stabilize variance and
approximate normality.
7. The prediction of the response variable has higher uncertainty than the estimation of the mean
response.
, A. False
B. True
Rationale: Predictions for new observations include both estimation error and inherent random
error, increasing uncertainty.
8. Statistical inference for linear regression under normality relies on large sample size.
A. True
B. False
Rationale: With normal errors, inference is exact regardless of sample size; large samples matter
mainly for CLT approximations.
9. A nonlinear relationship between response and predictor cannot be modeled using regression.
A. True
B. False
Rationale: Polynomial and transformed predictors allow modeling of nonlinear patterns within a
linear framework.
10. Assumption of normality in linear regression is required for confidence intervals, prediction
intervals, and hypothesis testing.
A. False
B. True
Rationale: These inferential tools rely on normality of residuals for valid t and F distributions.
11. If the confidence interval for a regression coefficient contains zero, we interpret that the
coefficient is plausibly equal to zero.
