← Intuit Interview Insights

Intuit·Data Scientist·Technical Phone Screen·Senior

Senior
Jul 2026

Summary

Intuit data scientist interview that went deep on A/B testing statistics, like really deep. They wanted full derivations, not just 'run a z-test and move on,' so if you're fuzzy on where the z-test formula actually comes from, this round will find that out fast.

Questions Asked (4)

Q1

For an A/B test on a binary delivery outcome, which significance test would you use and why is it preferable to a t-test on proportions? Derive the test statistic from first principles starting from the Bernoulli distribution through to the normal approximation, and state the conditions under which that approximation holds.

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

This is where I spent most of my mental energy.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by stating that the two-proportion z-test is the standard choice for comparing binary outcomes in A/B tests, then derive it from the Bernoulli distribution using the Central Limit Theorem. Explain why it is preferable to a t-test on proportions, emphasizing the correct variance structure and the large-sample normal approximation. Finally, clearly state the conditions under which the normal approximation holds.

Pro tip: Mention that in practice, the z-test and t-test yield nearly identical results for large samples, but the z-test is theoretically correct because the variance of a Bernoulli is determined by its mean. Also, note that some practitioners use a t-test as a conservative approach when sample sizes are small, but it's not strictly necessary.

1. Identify the appropriate test

State that the two-proportion z-test is the correct significance test for comparing binary outcomes in an A/B test, as it directly models the Bernoulli distribution of each group.

2. Derive from Bernoulli to normal

Show that each observation is Bernoulli(p), the sum is Binomial(n, p), and by the Central Limit Theorem, the sample proportion is approximately normal for large n. Then construct the z-statistic using the difference in proportions and its standard error.

3. Explain why not a t-test

Highlight that a t-test assumes an unknown variance estimated from the data, but for Bernoulli data, the variance is a function of the mean (p(1-p)), so the z-test uses the correct variance without estimating an extra parameter. This makes the z-test more powerful and theoretically appropriate.

4. State conditions for normal approximation

List the conditions: large sample sizes (typically n*p ≥ 10 and n*(1-p) ≥ 10 for each group), independent observations, and random assignment. Mention that for small samples or rare events, exact tests like Fisher's exact test may be needed.

Key Points to Mention

  • Bernoulli distribution for each user's outcome (conversion or not)
  • Central Limit Theorem and normal approximation for sample proportions
  • Two-proportion z-test statistic formula: (p1 - p2) / sqrt(p*(1-p)*(1/n1 + 1/n2))
  • Difference between z-test and t-test: known vs. estimated variance, and why Bernoulli variance is known given p
  • Conditions: np and n(1-p) ≥ 10 for each group, independence, random assignment
  • Alternatives for small samples: Fisher's exact test or permutation tests

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

Q2

Using the large-sample A/B test data (control: 920/1000 delivered, variant: 880/1000 delivered), compute the p-value to three significant figures and a 95% confidence interval for the difference in proportions. Show your work.

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

Numerical computation question mid-interview is always a little stressful.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

First, compute the sample proportions for control and variant, then calculate the difference. Use the two-proportion z-test to compute the z-statistic and corresponding p-value, and construct a 95% confidence interval for the difference using the standard error. Show all formulas and intermediate calculations clearly.

Pro tip: Always state your hypotheses and check assumptions (e.g., sample size large enough for normal approximation) before diving into calculations. This demonstrates rigor and helps catch potential issues.

1. Define parameters and hypotheses

Let p_control and p_variant be the true delivery rates. State the null hypothesis H0: p_control - p_variant = 0 and alternative H1: p_control - p_variant ≠ 0 (two-sided).

2. Calculate sample proportions and difference

Compute p_hat_control = 920/1000 = 0.92, p_hat_variant = 880/1000 = 0.88, and the observed difference d = 0.92 - 0.88 = 0.04.

3. Compute standard error and test statistic

Under H0, use pooled proportion p_pool = (920+880)/(1000+1000) = 0.90. Standard error SE = sqrt(p_pool*(1-p_pool)*(1/1000 + 1/1000)) ≈ 0.013416. Z = d / SE ≈ 2.981.

4. Find p-value

For a two-sided test, p-value = 2 * P(Z > |2.981|) ≈ 2 * 0.00144 = 0.00288. To three significant figures, p-value = 0.00288.

5. Construct 95% confidence interval

Use unpooled SE for CI: SE_CI = sqrt(0.92*0.08/1000 + 0.88*0.12/1000) ≈ 0.013416. 95% CI = d ± 1.96*SE_CI = 0.04 ± 0.0263 = (0.0137, 0.0663).

Key Points to Mention

  • Two-proportion z-test for difference in proportions
  • Pooled vs unpooled standard error: pooled for hypothesis test, unpooled for confidence interval
  • Assumption of large sample size (np and n(1-p) > 5 or 10)
  • Two-sided p-value calculation
  • Interpretation of confidence interval: we are 95% confident the true difference lies within the interval
  • Practical significance vs statistical significance

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

Q3

A small-market rollout gives you much smaller sample sizes: 20 out of 25 delivered in control, 12 out of 18 in variant. How does your test choice change? Walk through whether you'd use chi-square, Fisher's exact, or a z-test, and either compute the exact p-value or explain precisely how you'd get it.

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

Fisher's exact was the right call and I knew it, small counts break the normal approximation.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by recognizing that with small sample sizes, the normal approximation for the z-test is unreliable, so you should choose between chi-square and Fisher's exact based on expected cell counts. Compute the exact p-value using Fisher's exact test (or explain how to compute it via hypergeometric probabilities) and discuss the trade-offs between the two tests.

Pro tip: Mention that Fisher's exact test is conservative but valid for small samples, while chi-square may be anti-conservative; also note that in practice, you might use a mid-p adjustment or simulation to balance power and validity.

1. Assess sample size and expected counts

Calculate the expected cell counts under the null hypothesis. With 20/25 and 12/18, the overall conversion rate is (20+12)/(25+18)=32/43≈0.744. Expected counts: control successes=25*0.744≈18.6, failures≈6.4; variant successes=18*0.744≈13.4, failures≈4.6. All expected counts are ≥5, but some are close, and total N=43 is small.

2. Evaluate test assumptions

The z-test for proportions assumes normal approximation, which is questionable with small samples and expected counts near 5. Chi-square test also relies on the same approximation and may be inaccurate. Fisher's exact test does not rely on large-sample approximation and is exact for 2x2 tables.

3. Choose the appropriate test

Given the small sample size and the need for exact inference, Fisher's exact test is the most appropriate. Chi-square could be used but may yield inaccurate p-values; z-test is not recommended. If using chi-square, apply Yates' continuity correction, but still note its limitations.

4. Compute or explain the exact p-value

For Fisher's exact test, fix the margins (25,18,20,12) and compute the probability of observing the given table or more extreme using the hypergeometric distribution. The p-value is the sum of probabilities of tables with the same margins that are as or more extreme than the observed. Alternatively, use statistical software (e.g., scipy.stats.fisher_exact) to get the p-value.

5. Interpret and discuss trade-offs

Report the p-value and discuss its interpretation. Note that Fisher's exact test is conservative (higher p-values) and may lack power. Mention that with small samples, confidence intervals are wide, and practical significance should be considered alongside statistical significance.

Key Points to Mention

  • Expected cell counts and the rule of thumb (all expected counts ≥5) for chi-square validity.
  • Fisher's exact test is based on the hypergeometric distribution and does not rely on large-sample approximations.
  • The z-test for proportions is inappropriate for small samples due to poor normal approximation.
  • Yates' continuity correction for chi-square, but it may be overly conservative.
  • Exact p-value computation: sum of hypergeometric probabilities for tables as extreme or more extreme than observed.
  • Trade-offs: Fisher's exact is conservative but valid; chi-square may be anti-conservative; consider mid-p or simulation for better power.

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

Q4

When does a two-sample t-test produce essentially the same result as a z-test for proportions, and when does it break down? Give at least two concrete failure modes.

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

Conceptually I had this but my examples were a bit generic.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the equivalence conditions: large sample sizes, moderate proportions, and independent observations. Then explain that the two-sample t-test on binary data approximates the z-test for proportions when the normal approximation to the binomial holds. Finally, detail at least two failure modes where the approximation breaks down, such as small sample sizes or extreme proportions.

Pro tip: Emphasize that in practice, the z-test for proportions and the t-test on 0/1 data yield nearly identical results when np and n(1-p) exceed 5 or 10, but the t-test is more robust to variance estimation issues. However, always check assumptions; for small samples or rare events, exact methods like Fisher's exact test are preferable.

1. Clarify the tests

Define the two-sample t-test for proportions (applied to 0/1 data) and the z-test for proportions, noting they both compare two independent proportions.

2. State equivalence conditions

Explain that they produce essentially the same result when sample sizes are large enough for the normal approximation (e.g., np ≥ 5 and n(1-p) ≥ 5 for both groups) and proportions are not extreme.

3. Identify failure mode 1: small sample sizes

When sample sizes are small, the t-test's reliance on the t-distribution and pooled variance may not approximate the normal well, leading to inaccurate p-values and confidence intervals.

4. Identify failure mode 2: extreme proportions

When proportions are near 0 or 1, the normal approximation is poor, and the t-test can produce invalid results (e.g., negative variance estimates or p-values > 1).

5. Mention additional failure modes

Discuss other breakdowns: unequal sample sizes with unequal variances, dependence between observations, or when using the t-test without continuity correction.

Key Points to Mention

  • Normal approximation condition: np and n(1-p) ≥ 5 or 10 for both groups.
  • The t-test on 0/1 data is equivalent to a two-sample z-test when variances are known and sample sizes are large.
  • Failure mode: small sample sizes lead to inflated Type I error or low power.
  • Failure mode: extreme proportions (near 0 or 1) violate normality and can yield nonsensical results.
  • Alternative methods: Fisher's exact test, chi-square test with continuity correction, or Bayesian approaches.
  • In A/B testing, always check assumptions and consider using exact tests for rare events.

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