← Meta Interview Insights

Meta·Data Scientist·Technical Phone Screen·Senior

Senior
Jul 2026

Summary

Tough technical screen at Meta for a data scientist role, heavy on sequential testing theory. The questions were genuinely hard and felt more like a stats PhD oral exam than a typical DS interview.

Questions Asked (3)

Q1

Compare Pocock and O'Brien-Fleming alpha-spending functions for an experiment with two interim looks. What are the corresponding critical z-values under each approach?

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

This is where I started to sweat.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by defining the two alpha-spending functions and their trade-offs, then compute the critical z-values for two interim looks using standard software or approximations. Emphasize how Pocock uses constant boundaries while O'Brien-Fleming is conservative early and lenient later, and relate this to practical implications for A/B testing at Meta.

Pro tip: Mention that in practice, O'Brien-Fleming is often preferred for its early conservatism, which aligns with Meta's emphasis on minimizing false positives in early experiment stages, but be prepared to discuss scenarios where Pocock's constant boundaries might be advantageous for faster decision-making.

1. Define the spending functions

Explain that Pocock's alpha-spending function allocates equal alpha at each look, leading to constant critical boundaries, while O'Brien-Fleming spends very little alpha early and more later, resulting in decreasing critical boundaries.

2. Specify the number of looks and overall alpha

State that there are two interim looks plus a final analysis (total 3 looks) and assume a two-sided alpha of 0.05 for concreteness, unless otherwise specified.

3. Compute or recall critical z-values

For Pocock with 3 looks and two-sided alpha 0.05, the critical z-value is approximately 2.289 at each look. For O'Brien-Fleming, the critical z-values are approximately 3.471 at the first look, 2.454 at the second, and 2.004 at the final.

4. Discuss trade-offs and practical implications

Highlight that Pocock may stop early more easily but has a higher risk of false positive if early data are noisy, while O'Brien-Fleming requires stronger evidence early, preserving overall alpha and is often favored in confirmatory trials.

5. Relate to A/B testing context

Connect to Meta's experimentation culture: O'Brien-Fleming is common for its conservative early stopping, but Pocock might be used when rapid iteration is prioritized, though with caution about inflated false positives.

Key Points to Mention

  • Alpha-spending functions control Type I error rate across multiple looks.
  • Pocock: constant critical boundaries; O'Brien-Fleming: decreasing boundaries.
  • For two interim looks (3 total), Pocock z ≈ 2.289; O'Brien-Fleming z ≈ 3.471, 2.454, 2.004.
  • O'Brien-Fleming is more conservative early, reducing chance of false positive from early noise.
  • Pocock may allow earlier stopping but at cost of higher alpha spent early.
  • Choice depends on trial goals: O'Brien-Fleming for confirmatory, Pocock for exploratory or when early decisions are critical.

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

Q2

How do mixture-SPRT or e-values allow valid inference no matter how many times you peek at the data?

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

Honestly the question I felt best about.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by explaining the multiple testing problem in sequential analysis and why fixed-horizon p-values fail under continuous monitoring. Then introduce mixture-SPRT and e-values as solutions that maintain type I error control regardless of peeking, and discuss their practical implications for A/B testing at Meta.

Pro tip: Emphasize that e-values are not just a technical fix but enable a new paradigm where you can stop experiments early or continue as needed without inflating false positives, which is crucial for rapid experimentation at scale.

1. Frame the problem

Explain that in standard A/B testing, peeking at data repeatedly inflates the false positive rate because each look is a separate test. This is a major issue in tech companies where experiments are monitored continuously.

2. Introduce mixture-SPRT

Describe mixture-SPRT as a sequential test that combines likelihood ratios with a prior over effect sizes, allowing valid inference at any stopping time. It controls type I error even under optional stopping.

3. Define e-values

Explain that an e-value is a nonnegative random variable with expectation at most 1 under the null hypothesis. It provides evidence against the null, and can be combined across analyses via multiplication, preserving validity under peeking.

4. Connect to anytime-valid inference

Highlight that both methods yield anytime-valid p-values or confidence sequences, meaning error guarantees hold uniformly over time, not just at a fixed sample size.

5. Discuss practical implications

Relate to Meta's experimentation culture: faster decisions, reduced opportunity cost, and the ability to monitor continuously without sacrificing statistical rigor. Mention trade-offs like wider confidence intervals or need for effect size priors.

Key Points to Mention

  • Multiple testing problem and alpha spending
  • Optional stopping and type I error inflation
  • Likelihood ratios and martingale properties
  • E-values as conservative p-values and their combination
  • Anytime-valid confidence sequences
  • Trade-offs: power, sample size, and prior specification

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

Q3

If you are running sequential monitoring across multiple metrics simultaneously, how do you keep overall error rates under control?

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

Blanked for a second here.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by acknowledging the multiple testing problem and the need to control the family-wise error rate (FWER) or false discovery rate (FDR). Then describe specific methods like alpha spending functions, group sequential designs, or error control procedures (e.g., Bonferroni, Holm, Benjamini-Hochberg) and how they apply to sequential monitoring. Finally, discuss practical trade-offs and how to choose the right approach based on the number of metrics, correlation, and business impact.

Pro tip: Emphasize that in practice, you often prioritize a primary metric with strict FWER control while using FDR control for secondary metrics, and mention that sequential testing requires adjusting alpha at each look to maintain overall error rates.

1. Define the error control goal

Clarify whether you need to control the family-wise error rate (FWER) or false discovery rate (FDR) across all metrics and interim looks. This depends on the cost of false positives and the number of metrics.

2. Choose a multiple testing correction method

Select a method such as Bonferroni, Holm, or Benjamini-Hochberg to adjust p-values or thresholds for the number of metrics. Consider correlations between metrics and use methods like the Benjamini-Yekutieli for dependent tests if needed.

3. Incorporate sequential monitoring adjustments

Use alpha spending functions (e.g., O'Brien-Fleming, Pocock) or group sequential boundaries to control error rates at each interim analysis. Combine with multiple testing corrections by adjusting the overall alpha or using a gatekeeping procedure.

4. Monitor and adjust in real-time

Implement the chosen design with pre-specified stopping rules and adjust thresholds dynamically as data accumulates. Use software or scripts to compute adjusted p-values and confidence intervals at each look.

5. Validate and communicate results

After the experiment, report adjusted p-values and effect sizes, and explain how error rates were controlled. Conduct sensitivity analyses to ensure robustness.

Key Points to Mention

  • Family-wise error rate (FWER) vs. false discovery rate (FDR) and when to use each
  • Alpha spending functions (e.g., O'Brien-Fleming, Pocock) for sequential testing
  • Multiple testing correction methods: Bonferroni, Holm, Benjamini-Hochberg, and their assumptions
  • Group sequential designs and interim analysis boundaries
  • Correlation between metrics and methods to handle dependence (e.g., Benjamini-Yekutieli, multivariate sequential tests)
  • Practical trade-offs: power, sample size, and business impact of false positives vs. false negatives

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