Least significant difference (LSD) method MCQs With Answer provides B. Pharm students a focused review of the LSD post-hoc test used after ANOVA for pairwise comparisons. This introduction covers key concepts such as pooled error variance, MSE, t-critical values, significance level (alpha), protection by overall F-test, and practical applications in pharmacology, bioassay, formulation studies and experimental design. Understanding assumptions, limitations (inflated Type I error), calculation formula and interpretation of results prepares students for data analysis in drug research. These targeted MCQs reinforce theory, calculations, software output interpretation and decision-making for laboratory studies. Now let’s test your knowledge with 30 MCQs on this topic.
Q1. What is the primary purpose of the Least Significant Difference (LSD) method?
- To test for normality of residuals
- To perform pairwise comparisons after a significant ANOVA
- To replace ANOVA for overall group differences
- To estimate sample size for experiments
Correct Answer: To perform pairwise comparisons after a significant ANOVA
Q2. Which statistic is central to the LSD formula for pairwise mean comparisons?
- Mean square between groups (MSB)
- Mean square error (MSE)
- Coefficient of variation (CV)
- Standard error of the mean (SEM) only
Correct Answer: Mean square error (MSE)
Q3. The formula for LSD (two-sample equal n) commonly includes which of these components?
- t-critical × sqrt(2 × MSE / n)
- z-critical × sqrt(MSB / n)
- F-critical × sqrt(MSE / 2n)
- Chi-square × sqrt(MSE / n)
Correct Answer: t-critical × sqrt(2 × MSE / n)
Q4. When is it appropriate to apply Fisher’s LSD without adjustment?
- Always, regardless of ANOVA outcome
- Only after a significant overall F-test in ANOVA
- Only for non-parametric data
- When sample sizes vary widely
Correct Answer: Only after a significant overall F-test in ANOVA
Q5. Which assumption is NOT required for valid LSD comparisons?
- Independence of observations
- Homogeneity of variances across groups
- Normality of residuals
- Equal group means before testing
Correct Answer: Equal group means before testing
Q6. What is a main limitation of the LSD method in multiple comparisons?
- It is overly conservative and misses true differences
- It inflates Type I error rate when many comparisons are made
- It requires extremely large sample sizes
- It only works for two-group experiments
Correct Answer: It inflates Type I error rate when many comparisons are made
Q7. In a pharmacology bioassay with four treatments (n per group = 6), which degrees of freedom are used to find t-critical for LSD?
- df = number of treatments – 1
- df = total sample size – number of treatments
- df = total sample size – 1
- df = number of pairwise comparisons
Correct Answer: df = total sample size – number of treatments
Q8. What does “protected LSD” refer to?
- Applying LSD only when overall ANOVA F-test is significant
- Using LSD without checking assumptions
- Using LSD with Bonferroni correction
- Applying LSD before running ANOVA
Correct Answer: Applying LSD only when overall ANOVA F-test is significant
Q9. How does unequal sample size (n) between groups affect the LSD calculation?
- It has no effect; formula is unchanged
- The pooled variance is not estimable
- The LSD uses sqrt(MSE × (1/n1 + 1/n2)) instead of 2/n
- LSD cannot be computed for unequal n
Correct Answer: The LSD uses sqrt(MSE × (1/n1 + 1/n2)) instead of 2/n
Q10. Which post-hoc test is more conservative than LSD for controlling family-wise error?
- Tukey’s HSD
- Unprotected LSD
- Pairwise t-tests without adjustment
- No post-hoc test is more conservative
Correct Answer: Tukey’s HSD
Q11. In LSD, the t-critical value is taken from which distribution?
- Chi-square distribution
- t-distribution with pooled error df
- F-distribution with (k-1, N-k) df
- Normal distribution (z)
Correct Answer: t-distribution with pooled error df
Q12. If MSE = 4, n = 5 per group, and t-critical = 2.571, what is the LSD value for pairwise means?
- 2.571 × sqrt(2 × 4 / 5) = 3.646
- 2.571 × sqrt(4 / 5) = 2.304
- 2.571 × sqrt(2 / 5) = 1.628
- 2.571 × sqrt(8 / 5) = 3.646 (duplicate)
Correct Answer: 2.571 × sqrt(2 × 4 / 5) = 3.646
Q13. In B. Pharm research, a significant LSD difference between formulations indicates what?
- The formulations are clinically equivalent
- There is a statistically significant difference in the measured response
- The experiment must be repeated due to Type II error
- No conclusion can be drawn without Bonferroni correction
Correct Answer: There is a statistically significant difference in the measured response
Q14. Which software output element is essential to compute LSD manually?
- Total sum of squares only
- Mean square error (MSE) and error degrees of freedom
- Group means only
- R-squared value only
Correct Answer: Mean square error (MSE) and error degrees of freedom
Q15. When comparing many treatments, which strategy reduces the risk of inflated Type I error compared to unadjusted LSD?
- Performing only the largest pairwise differences
- Using more liberal alpha (e.g., 0.10)
- Using more conservative post-hoc tests (Tukey, Bonferroni)
- Reducing sample size
Correct Answer: Using more conservative post-hoc tests (Tukey, Bonferroni)
Q16. The LSD method assumes homogeneity of variances. Which test can check this assumption?
- Shapiro-Wilk test
- Levene’s test
- Kaplan-Meier test
- Cochran’s Q for paired data
Correct Answer: Levene’s test
Q17. If ANOVA F-test is not significant, what is the recommended action regarding LSD comparisons?
- Proceed with LSD regardless
- Avoid LSD pairwise tests as they are not protected
- Increase alpha for LSD
- Switch to non-parametric LSD
Correct Answer: Avoid LSD pairwise tests as they are not protected
Q18. For which experimental design is LSD applicable for pairwise comparisons?
- Only completely randomized designs
- Any design analyzed by ANOVA if assumptions met (CRD, RBD, factorial)
- Only paired t-test designs
- Only non-parametric designs
Correct Answer: Any design analyzed by ANOVA if assumptions met (CRD, RBD, factorial)
Q19. Which term describes the pooled estimate of within-group variability used in LSD?
- Between-group variance
- Mean square error (MSE)
- Total variance
- Group variance
Correct Answer: Mean square error (MSE)
Q20. In a bioassay comparing three dose levels, which pairwise result would be declared significant using LSD?
- Difference between means = 0.5 when LSD = 0.8
- Difference between means = 1.2 when LSD = 1.0
- Difference between means = 0.9 when LSD = 1.0
- Difference between means = 0.2 when LSD = 0.15
Correct Answer: Difference between means = 1.2 when LSD = 1.0
Q21. Which phrase best describes Fisher’s LSD compared to family-wise methods?
- More conservative and reduces Type I error
- Less conservative and more powerful but risks inflated Type I error
- Identical in error control to Bonferroni
- Not based on pooled variance
Correct Answer: Less conservative and more powerful but risks inflated Type I error
Q22. In reporting LSD results in a B. Pharm article, which information should be included?
- MSE, df for error, t-critical, LSD value and which comparisons were significant
- Only p-values without test statistics
- Only raw means without error estimates
- Only sample sizes without ANOVA details
Correct Answer: MSE, df for error, t-critical, LSD value and which comparisons were significant
Q23. If researchers apply LSD to exploratory data with many comparisons, what is a common recommendation?
- Report results as confirmatory without caution
- Use LSD and ignore family-wise error
- Label conclusions as exploratory and consider adjustment or replication
- Always prefer LSD over all other methods
Correct Answer: Label conclusions as exploratory and consider adjustment or replication
Q24. Which outcome from an ANOVA justifies conducting LSD tests?
- Non-significant F with p > 0.05
- Significant F with p < alpha
- High R-squared value only
- Significant Levene’s test
Correct Answer: Significant F with p < alpha
Q25. How does the number of treatments (k) affect the number of pairwise comparisons for LSD?
- Number of comparisons = k
- Number of comparisons = k – 1
- Number of comparisons = k(k – 1)/2
- Number of comparisons = 2^k
Correct Answer: Number of comparisons = k(k – 1)/2
Q26. In an ANOVA table, which value is directly used to compute MSE?
- Error sum of squares divided by error df
- Total sum of squares divided by total df
- Between treatments sum of squares divided by treatments df
- R-squared multiplied by variance
Correct Answer: Error sum of squares divided by error df
Q27. When comparing two means with unequal variances, the standard LSD is inappropriate. What is a remedy?
- Ignore variance inequality and proceed
- Use Welch’s t-test or adjust degrees of freedom for unequal variances
- Always increase sample size to equalize variances
- Use LSD with pooled variance regardless
Correct Answer: Use Welch’s t-test or adjust degrees of freedom for unequal variances
Q28. Which of the following best describes the role of LSD in factorial experiments?
- LSD cannot be used in factorial experiments
- LSD can compare main effect levels or interaction cell means after appropriate ANOVA
- LSD replaces interaction terms in the model
- LSD is used to test block effects only
Correct Answer: LSD can compare main effect levels or interaction cell means after appropriate ANOVA
Q29. If the pooled MSE decreases while other factors remain constant, what happens to the LSD value?
- LSD increases
- LSD decreases
- LSD remains unchanged
- LSD becomes infinite
Correct Answer: LSD decreases
Q30. Which interpretation is correct if a pairwise mean difference exceeds the LSD threshold?
- The difference is statistically significant at the chosen alpha
- The difference is not statistically significant
- The F-test must be non-significant
- No interpretation is possible without Bonferroni correction
Correct Answer: The difference is statistically significant at the chosen alpha
