CLC Research Methods & Evidence-Based Practice 3 — Questions and Answers
Question 1: A laboratory scientist wants to determine if a new hemoglobin analyzer agrees with the reference method. Which statistical approach is most appropriate?
- Pearson correlation coefficient alone
- Bland-Altman analysis (Correct answer)
- Chi-square test
- Kaplan-Meier survival analysis
Correct answer: Bland-Altman analysis
Bland-Altman analysis assesses agreement between two measurement methods by plotting the difference against the mean of paired measurements.
Question 2: In a crossover study evaluating two specimen collection protocols, the primary advantage over a parallel-group design is:
- Elimination of selection bias through randomization
- Each participant serves as their own control, reducing inter-subject variability (Correct answer)
- Larger sample sizes can be enrolled
- Blinding is easier to maintain
Correct answer: Each participant serves as their own control, reducing inter-subject variability
Crossover designs use each participant as their own control, which reduces variability due to between-subject differences and often requires fewer participants.
Question 3: Which level of evidence, according to the Oxford Centre for Evidence-Based Medicine hierarchy, provides the strongest support for a clinical laboratory guideline?
- Expert opinion and editorials
- Case series without controls
- Systematic reviews of randomized controlled trials (Correct answer)
- Individual cohort studies
Correct answer: Systematic reviews of randomized controlled trials
Systematic reviews of high-quality RCTs represent Level 1 evidence, the highest tier in evidence hierarchies used for guideline development.
Question 4: A laboratory consultant conducts a study and finds p=0.04 for the primary outcome. A colleague points out the confidence interval crosses the line of no effect. Which statement is correct?
- The p-value always takes precedence over the confidence interval
- There is a contradiction; the CI and p-value should be concordant when calculated from the same data (Correct answer)
- The confidence interval is irrelevant if p<0.05
- The study results are statistically and clinically significant
Correct answer: There is a contradiction; the CI and p-value should be concordant when calculated from the same data
The p-value and confidence interval are mathematically linked; if the 95% CI crosses the null value, the p-value should be >0.05, indicating a computational error.
Question 5: A pre-analytical variable study uses stratified random sampling of 500 specimens. What is the purpose of stratification?
- To increase sample size
- To ensure subgroups of interest are adequately represented (Correct answer)
- To eliminate confounding variables
- To blind researchers to specimen identity
Correct answer: To ensure subgroups of interest are adequately represented
Stratified sampling divides the population into subgroups (strata) and samples from each, ensuring proportional or equal representation of key characteristics.
Question 6: A Funnel plot used in meta-analysis is primarily designed to detect:
- Heterogeneity between studies
- Publication bias (Correct answer)
- Confounding variables
- Random measurement error
Correct answer: Publication bias
Funnel plots assess publication bias by checking whether small studies with negative results are underrepresented, which would cause asymmetry in the plot.
Question 7: A laboratory evaluates a new troponin assay using a verification study with 120 samples. Which CLSI guideline is most applicable for this validation process?
- CLSI EP05 (Precision)
- CLSI EP09 (Method Comparison and Bias Estimation)
- CLSI EP15 (User Verification of Precision and Estimation of Bias) (Correct answer)
- CLSI EP06 (Linearity)
Correct answer: CLSI EP15 (User Verification of Precision and Estimation of Bias)
CLSI EP15 provides protocols specifically for user verification of precision and bias estimation when confirming a manufacturer's performance claims.
CLC Research Methods & Evidence-Based Practice 3
A laboratory scientist wants to determine if a new hemoglobin analyzer agrees with the reference method.
Which statistical approach is most appropriate?