"Excludability Assumption: Key to Accurate A/B Testing"

Excludability Assumption for One-Sided Non-Compliance

The excludability assumption for one-sided non-compliance is a statistical assumption that states that the treatment group and control group are comparable, except for the fact that some individuals in the treatment group do not comply with the treatment. This means that the non-compliant individuals can be excluded from the analysis without biasing the results.

This assumption is important in A/B testing on web and mobile because it allows for accurate analysis of the impact of a particular change or intervention. If the assumption is not met, the results of the test may be biased and not accurately reflect the true impact of the change.

Impact of Excludability Assumption on A/B Testing Web Mobile
Positive Impact Accurate analysis of the impact of a change on website metrics such as conversion rate, bounce rate, and time on page. Accurate analysis of the impact of a change on mobile app metrics such as retention rate, session length, and in-app purchases.
Negative Impact Biased results that do not accurately reflect the impact of the change, leading to incorrect decisions and wasted resources. Biased results that do not accurately reflect the impact of the change, leading to incorrect decisions and wasted resources.

Topic

  • Home

    Read more