What is the purpose of sensitivity analysis?

• A. Determine statistical significance
• B. Assess robustness
• C. Estimate effect size
• D. Identify data entry errors

Answer: (B)

Sensitivity analysis asks how results change when you tweak key assumptions, inputs, or methods. The aim is to see whether the main conclusions are robust across a range of plausible scenarios. If the study’s findings stay similar when you vary things like the method of handling missing data, inclusion of certain covariates, or the presence of outliers, you gain confidence that the results aren’t driven by a specific modeling choice or a particular subset of data. That focus on robustness is why this option is the best fit.

It’s not primarily about judging statistical significance, estimating a single effect size under one model, or identifying data entry errors. Those tasks are handled by other parts of the analysis or quality checks. Sensitivity analysis, by contrast, foregrounds how conclusions might shift with reasonable changes, highlighting uncertainty due to analytic decisions and data limitations.