← Meta Interview Insights

Meta·Data Scientist·Onsite - Multi Round·Senior

Senior
May 2026

Summary

Meta onsite for a Data Scientist role, this round was heavy on stats and experimentation theory. The interviewer clearly wanted to see whether you actually understand the math behind A/B testing or just know the buzzwords.

Questions Asked (1)

Q1

What's the difference between Type I and Type II errors in A/B testing, how would you decide on acceptable thresholds for each, and if your metric distribution is heavily skewed with outliers, how do you estimate treatment lift and build a confidence interval?

A/B Testing & ExperimentationProduct Analytics & Metrics
Author's notes

This is two questions stapled together and the second half is where people fall apart.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clearly defining Type I and Type II errors and their implications in A/B testing, then discuss how to set thresholds based on business costs and statistical power. For skewed distributions, explain robust methods like bootstrapping or transformations to estimate lift and confidence intervals.

Pro tip: Emphasize that thresholds should be driven by the relative cost of false positives vs. false negatives, and that for skewed metrics, bootstrapping is often more reliable than parametric methods because it doesn't assume normality.

1. Define Type I and Type II Errors

Explain that Type I error (false positive) is concluding a difference exists when it doesn't, while Type II error (false negative) is failing to detect a real difference. Relate these to significance level (α) and power (1-β).

2. Determine Acceptable Thresholds

Discuss how to set α and β based on business context: consider the cost of false positives (e.g., rolling out a bad feature) vs. false negatives (e.g., missing a good feature). Typical α is 0.05, but may be stricter for high-risk changes; power is often 0.8.

3. Address Skewed Distributions and Outliers

Acknowledge that skewed metrics violate normality assumptions, so standard t-tests may be invalid. Mention robust alternatives like bootstrapping, permutation tests, or transformations (e.g., log) to estimate lift and confidence intervals.

4. Estimate Lift and Confidence Interval

Describe using bootstrapping: resample with replacement, compute the metric difference (lift) for each resample, and derive the confidence interval from the bootstrap distribution. Alternatively, use trimmed means or quantile-based methods.

5. Validate and Interpret

Check that the chosen method is appropriate (e.g., bootstrap converges) and interpret the confidence interval in business terms, noting that outliers may still influence results and considering robust metrics like median or trimmed mean.

Key Points to Mention

  • Type I error rate (α) and Type II error rate (β), and their relationship to statistical power.
  • Trade-offs in setting thresholds: business impact, sample size, and multiple testing corrections.
  • Limitations of parametric tests (e.g., t-test) with skewed data and outliers.
  • Bootstrapping as a non-parametric method for confidence intervals and lift estimation.
  • Alternative approaches: permutation tests, log transformations, trimmed means, or quantile regression.
  • Importance of pre-registering analysis plans and sensitivity checks.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.