← Meta Interview Insights

Meta·Data Scientist·Technical Phone Screen·Senior

Senior
May 2026

Summary

Meta data scientist interview with a heavy stats/experimentation focus. The whole thing was basically one long scenario about a noisy A/B test, broken into four parts that got progressively more uncomfortable to answer under pressure.

Questions Asked (4)

Q1

You ran a two-sample t-test with 80% power and observed a positive effect, but the result wasn't statistically significant. Why can that happen even with a large sample?

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

I knew the textbook answer here but fumbled the delivery a bit.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying that 80% power means an 80% chance of detecting a true effect of a pre-specified size, so 20% of the time you'll miss it even when the effect is real. Then explain that a non-significant result with a positive observed effect can occur due to sampling variability, effect size smaller than assumed, or inflated variance. Finally, discuss practical implications and next steps.

Pro tip: Emphasize that power is about the true effect size, not the observed one; a positive but non-significant result often means the observed effect is smaller than the minimum detectable effect, which is common in practice. Also, mention that peeking or multiple testing can inflate false negatives if not adjusted.

1. Define Power and Significance

Briefly explain that power is the probability of rejecting the null when a true effect of a specific size exists, and significance is about the observed data. Clarify that they are related but distinct.

2. Explain Sampling Variability

Discuss how random sampling can produce a positive effect estimate that is not statistically significant due to chance, even with high power. Mention that the observed effect is a point estimate with uncertainty.

3. Consider Effect Size and Variance

Point out that if the true effect is smaller than the assumed effect size used in power calculation, power drops. Also, if variance is higher than expected, the test is less sensitive.

4. Address Multiple Comparisons and Peeking

If applicable, note that looking at data multiple times or testing many metrics can increase the chance of false negatives or require adjustments that reduce power.

5. Discuss Practical Implications

Suggest that a non-significant result doesn't prove no effect; it may warrant further investigation, larger sample, or consideration of practical significance.

Key Points to Mention

  • Power is the probability of detecting an effect of a given size, not a guarantee.
  • The observed effect size can be positive but not significant due to sampling error.
  • If the true effect is smaller than the minimum detectable effect, power is lower than planned.
  • Higher-than-expected variance reduces the effective power.
  • Multiple testing or peeking can affect error rates and power.
  • A non-significant result does not mean the effect is zero; it may be inconclusive.

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

Q2

Should you calculate post-hoc power after seeing the results? If not, what should you report instead and why?

A/B Testing & ExperimentationTechnical Trade-offs
Author's notes

This is the one I felt best about.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clearly stating that post-hoc power is not informative and should not be calculated after seeing results. Explain that it is a deterministic function of the p-value and adds no new information. Then, describe what to report instead, such as confidence intervals, effect sizes, and sensitivity analyses, and why these are more useful for interpreting the results.

Pro tip: Emphasize that post-hoc power can be misleading because it conflates effect size and sample size, and instead focus on the practical significance of the observed effect. Mention that reporting confidence intervals allows stakeholders to assess the range of plausible effects, which is more actionable for decision-making.

1. State the problem with post-hoc power

Explain that post-hoc power is calculated using the observed effect size and is therefore a direct function of the p-value. It does not provide additional information about the true power of the test.

2. Explain why it's not useful

Highlight that post-hoc power is often misinterpreted and can lead to erroneous conclusions, such as dismissing a non-significant result as underpowered when the observed effect is small.

3. Recommend alternative reporting

Suggest reporting confidence intervals for the effect size, which convey the precision of the estimate and the range of plausible effects. Also, report the observed effect size and its practical significance.

4. Discuss sensitivity analysis

Mention that if power is a concern, a sensitivity analysis can be done to determine the minimum detectable effect given the sample size, but this should be done a priori or as part of planning, not post-hoc.

5. Connect to decision-making

Emphasize that the goal is to inform decisions, so focus on the effect size and its uncertainty rather than on power. This helps stakeholders understand the practical implications of the results.

Key Points to Mention

  • Post-hoc power is a deterministic function of the p-value and observed effect size.
  • It does not provide new information beyond the p-value.
  • Confidence intervals provide a range of plausible effect sizes and are more informative.
  • Effect sizes should be interpreted in the context of practical significance.
  • Sensitivity analysis or minimum detectable effect should be planned a priori.
  • Post-hoc power can lead to the 'power approach' fallacy, where non-significant results are incorrectly attributed to low power.

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

Q3

Given a loss function where false positives are twice as costly as false negatives, how do you make a decision here and what are your next steps: extend the experiment, reduce variance, or stop?

A/B Testing & ExperimentationTechnical Trade-offsAdaptability & Ambiguity
Author's notes

Honestly the hardest part of the whole interview.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by translating the asymmetric loss into a concrete decision rule: compare the expected cost of shipping versus not shipping, incorporating the effect size, confidence interval, and the 2:1 cost ratio. Then assess whether the current experiment provides enough evidence to make a confident decision, and if not, determine whether extending the experiment or reducing variance is the most efficient path forward.

Pro tip: Frame your answer around expected loss minimization rather than just statistical significance, and explicitly state the threshold (e.g., posterior probability that the treatment is better) that would justify shipping given the cost asymmetry. This shows you think like a business-minded data scientist, not just a statistician.

1. Quantify the decision threshold

Convert the 2:1 cost ratio into a decision rule: ship if the expected loss from false positives is less than the expected loss from false negatives. For example, if the treatment effect is positive but uncertain, compute the posterior probability that the true effect exceeds zero and compare it to a threshold derived from the cost ratio.

2. Evaluate current evidence

Examine the observed effect size, confidence interval, and p-value. Determine whether the result is statistically significant and practically significant given the cost asymmetry. If the confidence interval is wide and includes both positive and negative effects, the evidence is inconclusive.

3. Assess decision risk

Calculate the expected cost of shipping if the treatment is actually harmful (false positive) versus not shipping if it is beneficial (false negative). Use the 2:1 ratio to weigh these costs. If the expected cost of a wrong decision is high, more evidence is needed.

4. Choose next step: extend, reduce variance, or stop

If the evidence is inconclusive and the potential gain from continuing is high, extend the experiment. If variance is the main issue and you can reduce it (e.g., by using CUPED, stratification, or more precise metrics), do so. If the expected cost of continuing outweighs the benefit, stop and make a decision based on current evidence.

5. Communicate and document

Clearly explain the decision, the rationale, and the trade-offs to stakeholders. Document the assumptions and the cost-benefit analysis for future reference.

Key Points to Mention

  • Asymmetric loss function and expected cost minimization
  • Bayesian decision theory or frequentist hypothesis testing with adjusted significance levels
  • Confidence intervals and effect size vs. statistical significance
  • Variance reduction techniques (CUPED, stratification, regression adjustment)
  • Sequential testing and peeking problem when extending experiments
  • Business impact and stakeholder communication

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

Q4

How would you redesign the experiment for the next iteration, specifically around the MDE, variance reduction techniques, and duration?

A/B Testing & ExperimentationProduct Analytics & MetricsTechnical Trade-offs
Author's notes

Ran through stratification and CUPED as variance reduction options, then talked about recalibrating the MDE based on what we now know the effect might actually be.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by diagnosing why the previous experiment fell short—whether it was underpowered, noisy, or too long—then propose targeted changes to MDE, variance reduction, and duration that balance statistical rigor with product velocity. Frame your redesign as a trade-off exercise, explicitly stating assumptions and how you'd validate them with a pilot or power analysis.

Pro tip: At Meta, speed matters: show you can reduce duration without sacrificing power by combining variance reduction (e.g., CUPED) with a slightly relaxed MDE, but always quantify the risk of false negatives and propose a guardrail metric to catch regressions.

1. Diagnose the previous experiment

Identify why the last iteration failed to detect an effect: was the MDE too large, variance too high, or duration too short? Use observed data to estimate actual variance and required sample size.

2. Re-evaluate the MDE

Determine the smallest effect size that is practically meaningful for the business. Balance statistical power (e.g., 80%) with the cost of longer experiments, and consider whether a larger MDE is acceptable for a quick directional read.

3. Apply variance reduction techniques

Propose methods like CUPED, stratification, or regression adjustment using pre-experiment covariates to reduce variance and increase sensitivity without increasing sample size.

4. Optimize duration and sample size

Recalculate required sample size and duration using the new MDE and reduced variance. Consider sequential testing or a fixed-horizon design, and account for novelty effects and weekly seasonality.

5. Validate and iterate

Run a pilot or simulation to validate assumptions, and set up monitoring for guardrail metrics. Be prepared to adjust MDE or duration if early results suggest the design is still underpowered.

Key Points to Mention

  • Minimum Detectable Effect (MDE) and its relationship to sample size, power, and business impact
  • Variance reduction techniques: CUPED, stratification, regression adjustment, and their assumptions
  • Trade-offs between experiment duration, traffic allocation, and statistical power
  • Sequential testing / peeking problem and how to handle it (e.g., alpha spending)
  • Novelty effects, primacy effects, and seasonality that can bias results
  • Guardrail metrics and how to balance sensitivity with false positive risk

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