STA 3032
Engineering Statistics
4.0 Credits
Final Exam Review (Qns & Ans)
2025
©2025
, Question 1:
In reliability analysis, which probability distribution is most widely
used to model time‑to‑failure data when the hazard rate may be
increasing, constant, or decreasing?
A. Exponential
B. Weibull
C. Lognormal
D. Uniform
Correct ANS: B. Weibull
Rationale:
The Weibull distribution is renowned for its flexibility—it can model
increasing, constant, or decreasing failure rates depending on the
value of its shape parameter, making it the standard choice in
reliability engineering.
---
Question 2:
In Bayesian inference for engineering statistics, using a conjugate
prior is advantageous because it:
A. Simplifies computations by ensuring the posterior is in the same
family as the prior
B. Eliminates uncertainty in parameter estimates
C. Requires fewer data points for accurate estimation
D. Automatically validates the model assumptions
Correct ANS: A. Simplifies computations by ensuring the
posterior is in the same family as the prior
Rationale:
A conjugate prior results in closed‑form solutions for the posterior
distribution, making the updating process computationally
straightforward and economical in practice.
---
©2025
, Question 3:
Which resampling method is most commonly used to approximate
the sampling distribution of a statistic when the underlying
distribution is unknown?
A. Cross‑validation
B. Jackknife
C. Bootstrapping
D. Permutation tests
Correct ANS: C. Bootstrapping
Rationale:
Bootstrapping relies on repeatedly resampling from the observed
data with replacement, offering a powerful non‑parametric
approach to estimate variability and confidence intervals without
assuming a specific underlying distribution.
---
Question 4:
In experimental design, the primary purpose of blocking is to:
A. Randomize the assignment of treatments
B. Reduce variability by grouping homogeneous experimental units
C. Increase the overall sample size
D. Enhance the resolution of factorial design
Correct ANS: B. Reduce variability by grouping homogeneous
experimental units
Rationale:
Blocking minimizes the impact of confounding variables by
ensuring that comparisons are made among similar experimental
units, thereby reducing within‑group variance.
---
Question 5:
In classical hypothesis testing, a Type I error is defined as:
A. Failing to reject a false null hypothesis
©2025
Engineering Statistics
4.0 Credits
Final Exam Review (Qns & Ans)
2025
©2025
, Question 1:
In reliability analysis, which probability distribution is most widely
used to model time‑to‑failure data when the hazard rate may be
increasing, constant, or decreasing?
A. Exponential
B. Weibull
C. Lognormal
D. Uniform
Correct ANS: B. Weibull
Rationale:
The Weibull distribution is renowned for its flexibility—it can model
increasing, constant, or decreasing failure rates depending on the
value of its shape parameter, making it the standard choice in
reliability engineering.
---
Question 2:
In Bayesian inference for engineering statistics, using a conjugate
prior is advantageous because it:
A. Simplifies computations by ensuring the posterior is in the same
family as the prior
B. Eliminates uncertainty in parameter estimates
C. Requires fewer data points for accurate estimation
D. Automatically validates the model assumptions
Correct ANS: A. Simplifies computations by ensuring the
posterior is in the same family as the prior
Rationale:
A conjugate prior results in closed‑form solutions for the posterior
distribution, making the updating process computationally
straightforward and economical in practice.
---
©2025
, Question 3:
Which resampling method is most commonly used to approximate
the sampling distribution of a statistic when the underlying
distribution is unknown?
A. Cross‑validation
B. Jackknife
C. Bootstrapping
D. Permutation tests
Correct ANS: C. Bootstrapping
Rationale:
Bootstrapping relies on repeatedly resampling from the observed
data with replacement, offering a powerful non‑parametric
approach to estimate variability and confidence intervals without
assuming a specific underlying distribution.
---
Question 4:
In experimental design, the primary purpose of blocking is to:
A. Randomize the assignment of treatments
B. Reduce variability by grouping homogeneous experimental units
C. Increase the overall sample size
D. Enhance the resolution of factorial design
Correct ANS: B. Reduce variability by grouping homogeneous
experimental units
Rationale:
Blocking minimizes the impact of confounding variables by
ensuring that comparisons are made among similar experimental
units, thereby reducing within‑group variance.
---
Question 5:
In classical hypothesis testing, a Type I error is defined as:
A. Failing to reject a false null hypothesis
©2025
