← Meta Interview Insights

Meta·Data Scientist·Technical Phone Screen·Senior

Senior
Apr 2026

Summary

Meta DS interview with a heavy stats and experimentation focus. The whole thing was essentially one long A/B testing case study with five sub-parts, which sounds manageable until you're mid-calculation and second-guessing your pooled proportion formula.

Questions Asked (5)

Q1

Two search models each served 100 users. Model A had 90 successes, Model B had 85. Using a two-sided test at alpha=0.05, state your null and alternative hypotheses, choose the right test, compute the test statistic and p-value, and conclude whether A outperforms B.

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

Went with a pooled two-proportion z-test which felt right.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clearly stating the null and alternative hypotheses for a two-proportion z-test, then compute the pooled proportion and standard error to get the z-statistic and p-value. Finally, compare the p-value to alpha=0.05 and conclude whether there is sufficient evidence that Model A outperforms Model B.

Pro tip: Mention that this is a one-sided test in disguise: although the question says 'two-sided test', the conclusion about A outperforming B requires a directional interpretation. Clarify that the two-sided p-value tests for any difference, but to claim A > B you need to check the direction of the effect and possibly halve the p-value (or use a one-sided test).

1. State hypotheses

Define null hypothesis H0: p_A = p_B (no difference in success rates) and alternative hypothesis H1: p_A ≠ p_B (two-sided). Optionally note that the directional alternative H1: p_A > p_B would be one-sided.

2. Choose the test

Select a two-proportion z-test for independent samples, as we are comparing two proportions from independent groups with large enough sample sizes (n=100 each).

3. Compute test statistic

Calculate the pooled proportion p_hat = (90+85)/(100+100) = 0.875. Then compute the standard error SE = sqrt(p_hat*(1-p_hat)*(1/100+1/100)) ≈ 0.0468. The z-statistic is (0.90 - 0.85)/SE ≈ 1.069.

4. Find p-value

For a two-sided test, p-value = 2 * P(Z > |z|) ≈ 2 * 0.1425 = 0.285. (Using standard normal table or software.)

5. Conclusion

Since p-value (0.285) > alpha (0.05), fail to reject the null hypothesis. There is not enough evidence to conclude that Model A outperforms Model B.

Key Points to Mention

  • Two-proportion z-test assumptions: independent samples, binary outcome, large sample size (np and n(1-p) > 5).
  • Pooled proportion calculation and its use in standard error under the null hypothesis.
  • Difference between two-sided and one-sided tests, and how the question's phrasing affects interpretation.
  • Effect size: the observed difference is 5 percentage points, but not statistically significant.
  • Practical significance vs statistical significance: even if significant, consider if 5% lift is meaningful for the product.
  • Confidence interval for the difference in proportions as an alternative way to assess the effect.

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

Q2

Compute a 95% confidence interval for the difference pA minus pB and explain what it means for practical significance, not just statistical significance.

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

The CI part was fine mechanically, unpooled SE this time, interval comes out something like (-0.03, 0.13).

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clearly stating the formula for the confidence interval of the difference between two proportions, then compute the interval using the provided sample data. After obtaining the interval, interpret it in terms of practical significance by discussing the magnitude of the difference and its relevance to business metrics, not just whether it excludes zero.

Pro tip: Always relate the confidence interval to the minimum detectable effect (MDE) or the business's threshold for meaningful impact; this shows you understand that statistical significance alone doesn't guarantee practical value.

1. State the formula and assumptions

Write down the formula for the confidence interval of the difference between two independent proportions: (pA - pB) ± z* sqrt( pA(1-pA)/nA + pB(1-pB)/nB ). Mention that this assumes large sample sizes and independent samples.

2. Compute the interval

Plug in the sample proportions and sizes, calculate the standard error, and find the margin of error using the z-score for 95% confidence (1.96). Then compute the lower and upper bounds of the interval.

3. Interpret statistical significance

Check if the interval contains zero. If it does not, the difference is statistically significant at the 5% level. Explain what this means in terms of the null hypothesis.

4. Assess practical significance

Discuss whether the entire interval lies above or below a threshold that matters for the business. For example, if the lower bound is greater than the minimum detectable effect or a business-relevant lift, then the result has practical significance.

5. Consider context and limitations

Mention other factors like sample size, duration of the experiment, novelty effects, and potential biases that could affect the practical implications. Emphasize that practical significance depends on the specific product and goals.

Key Points to Mention

  • Formula for confidence interval of difference between two proportions
  • Assumptions: independence, large sample size, random assignment
  • Interpretation of confidence interval: range of plausible values for the true difference
  • Statistical significance vs practical significance: interval excluding zero vs interval exceeding a meaningful threshold
  • Minimum detectable effect (MDE) or business-relevant threshold
  • Contextual factors: cost of implementation, user impact, and strategic goals

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

Q3

What per-arm sample size do you need to detect a 5 percentage-point lift over an 85% baseline with 80% power at alpha=0.05? Walk through the formula and inputs.

A/B Testing & Experimentation
Author's notes

Blanked for a second on whether to use one-sided or two-sided z-values here.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the test design (two-sided, equal allocation) and then walk through the sample size formula for comparing two proportions. Plug in the baseline rate (0.85), the target rate (0.90), alpha=0.05, and power=0.80, and compute the required per-arm sample size. Finally, sanity-check the result and mention practical considerations like multiple testing or sequential analysis.

Pro tip: Meta often runs many experiments simultaneously, so mention that you would adjust for multiple comparisons (e.g., Bonferroni) or use sequential testing to control false positives, which can increase sample size. Also, note that the formula assumes no interference and independent observations, which may not hold in social networks.

1. Clarify assumptions and design

Confirm the test is two-sided with equal allocation, and that the metric is a proportion (e.g., conversion rate). State that you assume independent observations and no interference.

2. Define inputs and formula

Identify p1=0.85, p2=0.90, alpha=0.05 (two-sided), power=0.80. Use the standard formula for two proportions: n = (Z_{α/2} + Z_β)^2 * (p1(1-p1) + p2(1-p2)) / (p2-p1)^2.

3. Compute critical values and plug in

Z_{α/2}=1.96, Z_β=0.84. Compute pooled variance or use unpooled as above. Calculate numerator and denominator to get n per arm.

4. Calculate and round up

Perform the arithmetic: (1.96+0.84)^2 = 7.84; p1(1-p1)=0.1275, p2(1-p2)=0.09, sum=0.2175; difference=0.05, squared=0.0025. n = 7.84 * 0.2175 / 0.0025 ≈ 682. Round up to 683 per arm.

5. Sanity-check and discuss practical considerations

Verify the result is reasonable (e.g., ~680 per arm). Mention that real-world factors like multiple testing, sequential analysis, or non-compliance may require larger samples.

Key Points to Mention

  • Two-sided test at alpha=0.05 and power=80% (beta=0.20).
  • Use of Z-scores: 1.96 for alpha/2 and 0.84 for beta.
  • Formula for sample size comparing two proportions.
  • Plug in p1=0.85, p2=0.90, and compute n per arm.
  • Result: approximately 683 per arm (rounded up).
  • Practical adjustments: multiple comparisons, sequential testing, or network effects.

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

Q4

You're running these two models across 10 independent intents simultaneously. Apply a Bonferroni correction and explain whether your conclusions change.

A/B Testing & Experimentation
Author's notes

Pretty quick answer: adjusted alpha becomes 0.05/10 = 0.005.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

First, clarify the setup: two models tested across 10 independent intents, likely with per-intent hypothesis tests. Then explain that Bonferroni correction adjusts the significance threshold to α/10 (e.g., 0.005 for α=0.05) to control the family-wise error rate. Finally, discuss whether the conclusions change by comparing the original p-values to the corrected threshold, and consider the trade-off between Type I error control and statistical power.

Pro tip: Mention that while Bonferroni is conservative, it's appropriate when false positives are costly; however, if the intents are truly independent and you're interested in each individually, you might use a less stringent method like Benjamini-Hochberg to control FDR, but be clear about the assumptions.

1. Clarify the experimental setup

Confirm that there are 10 independent hypothesis tests (one per intent) and that the goal is to compare two models. Identify the significance level (e.g., α=0.05) and whether the tests are two-sided.

2. Apply Bonferroni correction

Divide the original α by the number of tests (10) to get the corrected threshold (e.g., 0.005). Explain that this controls the family-wise error rate at 5%.

3. Re-evaluate conclusions per intent

Compare each intent's p-value to the corrected threshold. If a p-value was significant before but not after, the conclusion for that intent changes from significant to not significant.

4. Discuss trade-offs and alternatives

Acknowledge that Bonferroni reduces power and may miss true effects. Mention alternatives like Holm-Bonferroni or Benjamini-Hochberg if controlling false discovery rate is more appropriate.

5. Summarize impact on overall conclusions

State whether the overall conclusion about which model is better changes. If some intents lose significance, the evidence for a consistent winner may weaken, but if the effect is strong, it may persist.

Key Points to Mention

  • Family-wise error rate (FWER) and why it matters with multiple comparisons.
  • Bonferroni correction formula: α_corrected = α / m, where m is number of tests.
  • Impact on statistical power: Bonferroni is conservative, increasing Type II errors.
  • Independence assumption: Bonferroni does not require independence, but if tests are independent, other methods may be more powerful.
  • Alternatives: Holm-Bonferroni (step-down) and Benjamini-Hochberg (FDR control).
  • Practical significance vs. statistical significance: even if significance is lost, effect sizes and confidence intervals should be examined.

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

Q5

When would you prefer an exact test over a z-test, and when would you use a Bayesian comparison instead? What would you actually report in each case?

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

This was the most open-ended part and I think I did okay.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the decision context and data characteristics, then compare exact tests, z-tests, and Bayesian methods in terms of assumptions, robustness, and interpretability. Emphasize that the choice depends on sample size, variance assumptions, and the need for probabilistic statements, and always tie the reporting back to actionable metrics.

Pro tip: At Meta, where experiments run at massive scale, you'll rarely use exact tests for primary metrics, but you should know when they're appropriate for guardrail or rare-event metrics. Also, Bayesian methods are often used for decision-making under uncertainty, but you must clearly communicate the priors and decision criteria to stakeholders.

1. Clarify the decision context

Identify the goal: is it a frequentist hypothesis test, or do you need probability statements to support a business decision? Consider the cost of false positives/negatives and the need for interpretability.

2. Assess data and assumptions

Check sample size, distribution, variance homogeneity, and independence. Exact tests are preferred for small samples or when distributional assumptions are violated; z-tests require large samples and known variance.

3. Choose the method

Use exact tests (e.g., Fisher's exact, binomial) for small samples or rare events; z-tests for large-sample proportions/means with known variance; Bayesian methods when you want to incorporate prior information and directly quantify uncertainty.

4. Define what to report

For exact tests: p-value, confidence interval, effect size. For z-tests: z-statistic, p-value, confidence interval. For Bayesian: posterior distribution, credible interval, probability of superiority, and expected loss.

5. Connect to business impact

Translate the statistical results into practical terms: lift, revenue impact, and decision recommendations. Highlight any caveats or sensitivity analyses.

Key Points to Mention

  • Exact tests (e.g., Fisher's exact, binomial) are preferred when sample sizes are small or when the normal approximation is questionable, such as with rare events or sparse data.
  • z-tests are appropriate for large samples where the Central Limit Theorem applies and variance is known or well-estimated, common in large-scale A/B tests at Meta.
  • Bayesian comparisons are useful when you have prior information, want to directly compute the probability that one variant is better, or need to make decisions under uncertainty with a loss function.
  • Reporting for frequentist methods: p-values, confidence intervals, and effect sizes; for Bayesian: posterior probabilities, credible intervals, and expected loss.
  • Always consider the practical significance and business impact, not just statistical significance.
  • Be aware of multiple testing corrections and sequential testing in the context of A/B testing.

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