Optimization and factorial designs MCQs With Answer

Introduction: Optimization and factorial designs are core tools in Computer Aided Drug Development, enabling M.Pharm students to design efficient experiments, model responses, and find optimum formulation or process settings. This blog presents a focused set of multiple-choice questions covering factorial experiments, fractional factorials, response surface methodology (RSM), central composite and Box–Behnken designs, model building, interaction effects, aliasing, blocking, transformations, and multiresponse optimization. These MCQs emphasize practical interpretation, design selection, analysis, and validation strategies commonly applied in pharmaceutical formulation and process optimization. Use these questions to test and deepen your understanding of experimental design principles and their application in drug development workflows.

Q1. Which statement best describes a full factorial design?

• It varies one factor at a time while keeping others constant
• It tests all possible combinations of factor levels
• It only estimates main effects without interactions
• It is limited to two factors only

Correct Answer: It tests all possible combinations of factor levels

Q2. What is the primary advantage of factorial designs over one-factor-at-a-time (OFAT) experiments?

• They always require fewer experimental runs than OFAT
• They eliminate the need for randomization
• They allow estimation of interaction effects between factors
• They avoid the need for statistical analysis

Correct Answer: They allow estimation of interaction effects between factors

Q3. How many experimental runs are required for a full 2^3 factorial design without replicates?

• 6
• 8
• 9
• 12

Correct Answer: 8

Q4. In fractional factorial designs, what does the term “aliasing” refer to?

• Random variation of measurements across runs
• Confounding of two or more effects so they cannot be separated
• The use of center points to detect curvature
• A type of blocking to reduce nuisance variation

Correct Answer: Confounding of two or more effects so they cannot be separated

Q5. A Resolution IV fractional factorial design ensures which of the following properties?

• Main effects are aliased with other main effects
• Main effects are aliased with two-factor interactions
• Main effects are not confounded with any two-factor interactions
• All two-factor interactions are independently estimable with no confounding

Correct Answer: Main effects are not confounded with any two-factor interactions

Q6. Why are center points included in factorial or RSM experiments?

• To increase the number of factors that can be tested
• To provide an internal standard for instrument calibration
• To detect curvature (nonlinearity) and estimate pure error
• To replace the need for replicates entirely

Correct Answer: To detect curvature (nonlinearity) and estimate pure error

Q7. What is the main objective of Response Surface Methodology (RSM) in pharmaceutical development?

• To screen a very large number of factors quickly
• To build empirical models (often quadratic) to locate optimum factor settings
• To perform only two-level factorial screening
• To replace analytical method validation

Correct Answer: To build empirical models (often quadratic) to locate optimum factor settings

Q8. Which components are characteristic of a Central Composite Design (CCD)?

• Only two-level factorial points with no center points
• Factorial points, axial (star) points, and center points
• Only axial points arranged on a cube’s corners
• Randomized runs without any replicates

Correct Answer: Factorial points, axial (star) points, and center points

Q9. Which of the following is a distinctive feature of a Box–Behnken design?

• It includes all corner (extreme) combinations of factor levels
• It requires factors to be set at five levels each
• It uses combinations of midpoints of edges and center points, avoiding extreme corners
• It is identical to a full factorial design

Correct Answer: It uses combinations of midpoints of edges and center points, avoiding extreme corners

Q10. Which terms are included in a quadratic (second-order) response surface model?

• Only linear terms
• Linear and interaction terms, but no quadratic terms
• Linear, interaction, and quadratic (squared) terms
• Only quadratic terms without interactions

Correct Answer: Linear, interaction, and quadratic (squared) terms

Q11. What does the D-optimality criterion aim to maximize in experimental design?

• The number of experimental runs
• The determinant of the information matrix to minimize parameter estimate variance
• The amount of blocking in the design
• The number of center points

Correct Answer: The determinant of the information matrix to minimize parameter estimate variance

Q12. In multiresponse optimization, what is the desirability function approach?

• A method that discards all but the most important response
• A way to combine multiple responses into a single scale (0–1) objective for optimization
• An approach that requires each response to be optimized sequentially
• A procedure that does not allow weighting of responses

Correct Answer: A way to combine multiple responses into a single scale (0–1) objective for optimization

Q13. Which experimental design principle ensures that estimated factor effects are uncorrelated?

• Blocking
• Randomization
• Orthogonality
• Replication

Correct Answer: Orthogonality

Q14. When is a fractional factorial design most appropriate in formulation development?

• When the number of runs is unlimited and all interactions are of interest
• When only one factor is being studied
• When there are many factors but resources limit the number of experimental runs
• When a precise quadratic surface model is required for optimization

Correct Answer: When there are many factors but resources limit the number of experimental runs

Q15. What is the primary purpose of blocking in an experimental design?

• To increase the number of factor levels tested
• To intentionally confound main effects with interactions
• To reduce variability from known nuisance sources by grouping similar runs
• To force orthogonality between factors

Correct Answer: To reduce variability from known nuisance sources by grouping similar runs

Q16. Which diagnostic step is essential for validating a fitted RSM model?

• Only checking the R-squared value, no residual analysis needed
• Residual analysis, lack-of-fit test, and confirmation runs at predicted optimum
• Eliminating all interaction terms to simplify the model
• Using only center points to validate curvature

Correct Answer: Residual analysis, lack-of-fit test, and confirmation runs at predicted optimum

Q17. Which data transformation is commonly used to stabilize variance and approximate normality in response data?

• Logarithmic or Box–Cox transformation
• Converting continuous response into a categorical variable
• Random permutation of response values
• Subtracting the mean from each observation only

Correct Answer: Logarithmic or Box–Cox transformation

Q18. In ANOVA for factorial experiments, a significant interaction between two factors implies which of the following?

• The main effects are always nonsignificant
• The effect of one factor depends on the level of the other factor
• The experiment must be converted to OFAT to interpret results
• There is no need to examine main effects at all

Correct Answer: The effect of one factor depends on the level of the other factor

Q19. Which software package is widely used specifically for designing and analyzing RSM and DOE in pharmaceutical research?

• Microsoft Excel only
• Design-Expert
• ImageJ
• GraphPad Prism exclusively

Correct Answer: Design-Expert

Q20. Which combination of actions provides the strongest evidence that an experimentally optimized formulation is reliable?

• High model R-squared only
• Significant p-values for some coefficients with no replication
• Model diagnostics (residuals, lack-of-fit), validation experiments at predicted optimum, and robustness checks
• Single-run confirmation without randomization

Correct Answer: Model diagnostics (residuals, lack-of-fit), validation experiments at predicted optimum, and robustness checks

Download