Blocking and confounding in two-level factorial design are key concepts for B.Pharm students focused on formulation and process optimization. Blocking reduces nuisance variability by grouping runs by batch, operator, or equipment, while confounding (aliasing) occurs when treatment effects mix with block effects or interactions. In fractional two-level (2-level) factorials, generators and defining relations create specific alias patterns; understanding resolution, half-fractions, foldover strategies, orthogonality, and ANOVA interpretation is critical. Practical skills include building block-confounded designs, calculating aliases, identifying estimable contrasts, and selecting follow-up experiments. Now let’s test your knowledge with 30 MCQs on this topic.
Q1. What is the primary purpose of blocking in a two-level factorial experiment?
- To group experimental runs to control nuisance variability such as batch or operator effects
- To deliberately confound main effects with interactions
- To increase the number of factor levels
- To eliminate the need for randomization
Correct Answer: To group experimental runs to control nuisance variability such as batch or operator effects
Q2. How does confounding differ from blocking?
- Confounding mixes treatment effects with block or interaction effects; blocking arranges runs to control nuisance variation
- Blocking mixes treatment effects; confounding controls nuisance variation
- Blocking increases run size; confounding reduces the number of factors
- There is no difference; the terms are interchangeable
Correct Answer: Confounding mixes treatment effects with block or interaction effects; blocking arranges runs to control nuisance variation
Q3. In a two-level factorial design, what do the two levels typically represent?
- Quantitative levels such as 0 and 100 only
- Two categorical groups unrelated to factor effects
- High and low settings, commonly coded as +1 and −1
- Replicate and control conditions only
Correct Answer: High and low settings, commonly coded as +1 and −1
Q4. Why are fractional factorial designs used in pharmaceutical experiments?
- To explore all possible combinations regardless of cost
- To reduce the number of runs while accepting some confounding between effects
- To avoid any confounding of interactions
- To ensure all interactions are estimated with maximum precision
Correct Answer: To reduce the number of runs while accepting some confounding between effects
Q5. What is a generator in a fractional factorial design?
- An equation that defines how one factor equals the product of others, e.g., D = A B C
- The statistical software used to analyze the design
- A randomization schedule for runs
- The block label assigned to each run
Correct Answer: An equation that defines how one factor equals the product of others, e.g., D = A B C
Q6. What does the defining relation determine in a fractional factorial design?
- The alias structure among effects
- The sample size calculation method only
- How to randomize within blocks
- The cost of each experimental run
Correct Answer: The alias structure among effects
Q7. If a half-fraction of a 2^3 design is generated by D = A B C (so I = A B C D), which effect is A aliased with?
- A is aliased with B C D
- A is aliased with B
- A is aliased with D only
- A is aliased with B C
Correct Answer: A is aliased with B C D
Q8. What characterizes a Resolution III design?
- Main effects may be confounded with two-factor interactions
- Main effects are confounded only with other main effects
- Main effects are completely orthogonal to all interactions
- No effects are aliased
Correct Answer: Main effects may be confounded with two-factor interactions
Q9. Which effects are best to confound with block effects when blocking is required?
- High-order interactions that are assumed negligible
- Main effects that are critical to the study aim
- All measured responses equally
Correct Answer: High-order interactions that are assumed negligible
Q10. What is the primary purpose of performing a foldover in a fractional factorial study?
- To resolve aliasing and separate confounded effects
- To further reduce the number of runs
- To increase blocking across all factors
- To change factor levels from categorical to continuous
Correct Answer: To resolve aliasing and separate confounded effects
Q11. What does orthogonality between factor columns in the design matrix imply?
- Estimates of effects are uncorrelated and independently estimable
- Effects are necessarily confounded with blocks
- Randomization is not required
- There are no interactions in the system
Correct Answer: Estimates of effects are uncorrelated and independently estimable
Q12. How many runs are required for a half-fraction of a 2^4 design?
- 8 runs
- 4 runs
- 16 runs
- 2 runs
Correct Answer: 8 runs
Q13. In two-level coding (+1, −1), how is a main effect estimated from responses?
- Half the difference between the average responses at +1 and −1 levels
- The average of all responses irrespective of levels
- The sum of squares divided by the error
- By using only the runs at the +1 level
Correct Answer: Half the difference between the average responses at +1 and −1 levels
Q14. A complete 2^3 factorial has 8 runs. If these runs are split into two blocks, how many runs per block are there?
- 4 runs per block
- 2 runs per block
- 8 runs per block
- 1 run per block
Correct Answer: 4 runs per block
Q15. In blocking, block effects are commonly treated as which type of effect in analysis?
- Nuisance effects often modeled as random effects
- Primary fixed treatment effects of interest
- Interactions between factors only
- Ignored completely in the model
Correct Answer: Nuisance effects often modeled as random effects
Q16. In the defining relation I = A B C D, what does I represent?
- The identity element indicating the defining relation or identity generator
- The interaction between I and other factors
- The inverse of the design matrix
- The error term in ANOVA
Correct Answer: The identity element indicating the defining relation or identity generator
Q17. Under what assumption is confounding considered acceptable when choosing a fractional factorial?
- When higher-order interactions assumed negligible are the ones confounded with main effects
- When main effects are expected to be zero
- When no blocking is planned
- When all interactions are known to be large
Correct Answer: When higher-order interactions assumed negligible are the ones confounded with main effects
Q18. How many degrees of freedom does each main effect have in a two-level factorial analysis?
- 1 degree of freedom
- 2 degrees of freedom
- k degrees of freedom, where k is the number of factors
- Zero degrees of freedom
Correct Answer: 1 degree of freedom
Q19. Which analysis approach is appropriate for a blocked factorial experiment?
- ANOVA including block as a factor in the model
- Simple linear regression ignoring block
- Pairwise t-tests between all runs
- Principal component analysis only
Correct Answer: ANOVA including block as a factor in the model
Q20. If a main effect is discovered to be confounded with a block effect, what is a practical corrective measure?
- Use a foldover or redesign to separate the confounded effects
- Delete the block and proceed without it
- Assume the main effect is zero and continue
- Ignore replication entirely
Correct Answer: Use a foldover or redesign to separate the confounded effects
Q21. In a 2^4−1 half-fraction generated by D = A B C, which effect is A aliased with?
- A is aliased with B C D
- A is aliased with B only
- A is aliased with C only
- A is aliased with D only
Correct Answer: A is aliased with B C D
Q22. How is blocking commonly implemented in a fractional factorial design?
- By introducing a block-defining generator that is the product of selected treatment factor columns
- By increasing the number of runs to a full factorial only
- By removing one factor entirely from the design
- By randomizing blocks after analysis
Correct Answer: By introducing a block-defining generator that is the product of selected treatment factor columns
Q23. What does a Resolution IV design ensure?
- Main effects are not aliased with any two-factor interactions
- Main effects are aliased with other main effects
- Two-factor interactions are aliased with main effects
- No interactions are estimable
Correct Answer: Main effects are not aliased with any two-factor interactions
Q24. When planning blocks for a tablet formulation study, which factors should you prioritize keeping unconfounded?
- Critical formulation and process main effects expected to be important
- High-order interactions only
- All nuisance factors as main effects
- Only blocking factors, not treatment factors
Correct Answer: Critical formulation and process main effects expected to be important
Q25. In a design with defining relation I = A B C, what is AB aliased with?
- AB is aliased with C
- AB is aliased with A only
- AB is aliased with I only
- AB is aliased with D
Correct Answer: AB is aliased with C
Q26. If a detected significant effect could be either a main effect or an aliased two-factor interaction, what is the recommended next step?
- Perform a confirmatory experiment such as a foldover to de-alias the effect
- Assume it is the main effect and publish results immediately
- Ignore the effect as it cannot be trusted
- Only increase block size without additional runs
Correct Answer: Perform a confirmatory experiment such as a foldover to de-alias the effect
Q27. What does the sparsity-of-effects principle state?
- A system is typically dominated by a small number of significant effects
- All factorial effects are equally large
- No interactions are expected in well-designed experiments
- Only block effects matter, not treatment effects
Correct Answer: A system is typically dominated by a small number of significant effects
Q28. How many runs are in a half-fractional 2^5−1 design?
- 16 runs
- 32 runs
- 8 runs
- 4 runs
Correct Answer: 16 runs
Q29. If a main effect is confounded with a block effect, what is the consequence for estimating that main effect?
- It cannot be separated from block effect and cannot be uniquely estimated
- It becomes more precise and unbiased
- It will have more degrees of freedom than before
- It becomes orthogonal to all interactions
Correct Answer: It cannot be separated from block effect and cannot be uniquely estimated
Q30. Which practices help avoid accidentally confounding important effects with blocks?
- Plan generators to confound only high-order interactions, randomize within blocks, and include replication where possible
- Never randomize and always use the smallest number of runs
- Confound main effects deliberately to simplify analysis
- Ignore aliasing and proceed with full pooling of effects
Correct Answer: Plan generators to confound only high-order interactions, randomize within blocks, and include replication where possible
