Optimization concepts and parameters MCQs With Answer

Optimization Concepts and Parameters MCQs With Answer for M. Pharm (Modern Pharmaceutics MPH 103T)

Optimization is central to modern pharmaceutics, where formulation scientists systematically fine-tune materials and process variables to achieve target product profiles and robust performance. This quiz focuses on optimization concepts and parameters used in Quality by Design (QbD), experimental design, and response surface methodology. You will test your understanding of key ideas such as objective functions, design types (factorial, CCD, Box–Behnken, mixture), model adequacy metrics, desirability functions, and constraints. The questions are crafted for M. Pharm students to deepen practical insight into selecting designs, interpreting statistics, handling multi-response trade-offs, and defining design space. Use this set to strengthen conceptual clarity and prepare for advanced formulation development tasks.

Q1. In pharmaceutical optimization, the objective function typically represents:

• A set of fixed experimental conditions that cannot be altered
• A qualitative descriptor of product appearance
• A quantitative response (e.g., % drug released) to be maximized or minimized
• A regulatory guideline for clinical endpoints

Correct Answer: A quantitative response (e.g., % drug released) to be maximized or minimized

Q2. In Design of Experiments (DoE), the terms “factors” and “responses” most accurately refer to:

• Dependent variables and control limits, respectively
• Independent variables under investigator control and measured outcomes, respectively
• Noise variables and nuisance parameters, respectively
• Latent variables and derived variables, respectively

Correct Answer: Independent variables under investigator control and measured outcomes, respectively

Q3. Which design is most efficient for fitting a second-order model with three factors without requiring extreme axial points?

• Central Composite Design (CCD)
• Box–Behnken Design (BBD)
• Plackett–Burman Design
• Full factorial 2³ design

Correct Answer: Box–Behnken Design (BBD)

Q4. In a Central Composite Design with k factors, the rotatability condition is achieved by setting the axial distance (α) to:

• α = 2k
• α = √k
• α = (2^k)^(1/4)
• α = 1/k

Correct Answer: α = (2^k)^(1/4)

Q5. What is the fundamental constraint in mixture experiments used for formulation optimization?

• Each component must be present at equal levels
• The sum of component proportions equals 1 (or 100%)
• All factors must be orthogonal
• Total variance must be minimized

Correct Answer: The sum of component proportions equals 1 (or 100%)

Q6. In multi-response optimization using desirability functions, the overall desirability (D) is typically computed as:

• The arithmetic mean of individual desirabilities
• The minimum of individual desirabilities
• The geometric mean: D = (d1 × d2 × … × dm)^(1/m)
• The sum of squared desirabilities

Correct Answer: The geometric mean: D = (d1 × d2 × … × dm)^(1/m)

Q7. Which metric best reflects the predictive performance of an RSM model for new observations?

• Coefficient of determination (R²)
• Adjusted R²
• Predicted R² (based on PRESS)
• Mean of residuals

Correct Answer: Predicted R² (based on PRESS)

Q8. In a 2ᵏ factorial design augmented with center points, statistical evidence of curvature is obtained by:

• Comparing block means by t-test
• A curvature test based on the difference between center point mean and factorial point mean
• Computing only the main effects
• Using a normal probability plot of residuals

Correct Answer: A curvature test based on the difference between center point mean and factorial point mean

Q9. Which fractional factorial design resolution allows two-factor interactions to be estimated unconfounded with main effects and with each other?

• Resolution III
• Resolution IV
• Resolution V
• Resolution II

Correct Answer: Resolution V

Q10. For Taguchi’s “larger-the-better” quality characteristic, the signal-to-noise (S/N) ratio is calculated as:

• S/N = −10 log10[(1/n) Σ(1/yi²)]
• S/N = −10 log10[(1/n) Σ(yi)]
• S/N = −10 log10[(1/n) Σ(yi − ȳ)²]
• S/N = −10 log10[(1/n) Σ|yi|]

Correct Answer: S/N = −10 log10[(1/n) Σ(1/yi²)]

Q11. The method of steepest ascent (or descent) in RSM is primarily used to:

• Fit a second-order polynomial model directly
• Move iteratively in the direction of maximum increase (or decrease) of the response
• Reduce multicollinearity among factors
• Estimate lack-of-fit error

Correct Answer: Move iteratively in the direction of maximum increase (or decrease) of the response

Q12. A Box–Cox transformation is most appropriately applied to:

• Stabilize variance and improve normality of residuals
• Increase model degrees of freedom
• Alter factor levels to extended ranges
• Orthogonalize correlated factors

Correct Answer: Stabilize variance and improve normality of residuals

Q13. In canonical analysis of a second-order response surface, a saddle point is indicated when:

• All eigenvalues of the Hessian are positive
• All eigenvalues of the Hessian are negative
• Eigenvalues are of mixed signs
• The gradient vector is zero and the intercept is non-zero

Correct Answer: Eigenvalues are of mixed signs

Q14. Due to the mixture constraint, which term is typically omitted in canonical mixture models (e.g., Scheffé polynomials)?

• Quadratic cross-product terms
• Pure quadratic (squared) terms
• Intercept term
• All linear terms

Correct Answer: Intercept term

Q15. A D-optimal design selects experimental runs that:

• Minimize the total number of experiments irrespective of model
• Maximize the determinant of X′X to minimize parameter variance
• Ensure all factor levels are equally spaced
• Eliminate the need for model validation

Correct Answer: Maximize the determinant of X′X to minimize parameter variance

Q16. In desirability-based multi-response optimization, if any individual desirability (di) equals zero, the overall desirability (D) becomes:

• One
• Equal to the average of the remaining desirabilities
• Zero
• Undefined

Correct Answer: Zero

Q17. In QbD, an overlay plot on a design space is used to:

• Show only the main effects of factors on one response
• Display regions where all responses simultaneously meet specifications
• Estimate experimental error directly
• Replace confirmatory runs

Correct Answer: Display regions where all responses simultaneously meet specifications

Q18. A Critical Process Parameter (CPP) is best defined as:

• A parameter that is easy to measure and record
• A process parameter whose variability has no impact on product quality
• A process parameter that, when varied, can impact a Critical Quality Attribute and must be controlled
• A material attribute related only to raw materials

Correct Answer: A process parameter that, when varied, can impact a Critical Quality Attribute and must be controlled

Q19. Which diagnostic is most commonly used to detect multicollinearity in RSM models?

• Durbin–Watson statistic
• Shapiro–Wilk test
• Levene’s test
• Variance Inflation Factor (VIF)

Correct Answer: Variance Inflation Factor (VIF)

Q20. To quantify the probability of meeting multiple CQAs given realistic variability in factors and model error, an appropriate approach is:

• One-factor-at-a-time sensitivity testing
• Monte Carlo simulation using the fitted predictive model
• Normal probability plotting of residuals
• Principal component analysis of factors

Correct Answer: Monte Carlo simulation using the fitted predictive model

Download