Data Collection & Analysis Flashcards

In an intention-to-treat (ITT) analysis, how are participants analyzed?

Answer: In the group they were randomized to, regardless of adherence

ITT keeps everyone in their randomized group, which preserves the balance that randomization created.

Which analysis population includes only participants who followed the protocol closely enough with no major deviations?

Answer: Per-protocol population

The per-protocol population excludes participants with major deviations so it can estimate the effect under ideal adherence.

The 95% confidence interval for the difference in means between two groups is -1.2 to 3.5. What can be concluded at the 0.05 level?

Answer: The difference is not statistically significant because the interval includes zero

If a confidence interval for a difference includes zero, no difference is a plausible result, so it is not significant at alpha = 0.05.

What is the main reason for blinding outcome assessors in a clinical study?

Answer: To reduce measurement bias in outcome assessment

Assessors who do not know treatment assignments cannot let expectations influence how they rate outcomes.

Which group usually reviews unblinded interim data to decide whether a trial should stop early for safety or efficacy?

Answer: Data and Safety Monitoring Board

An independent DSMB reviews unblinded accumulating data and recommends whether to continue, modify, or stop the trial.

When should a study's primary endpoint be defined?

Answer: Before data collection, in the protocol and statistical analysis plan

Defining the primary endpoint in advance prevents selective reporting and data-driven outcome switching.

A correlation coefficient of r = -0.85 between exercise hours and resting heart rate indicates what?

Answer: A strong negative relationship

Values near -1 indicate a strong inverse linear relationship: as one variable rises, the other falls.