← Capital One Interview Insights

Capital One·Data Scientist·Technical Phone Screen·Intermediate

Intermediate
May 2026

Summary

Capital One data scientist interview that was basically a stats exam in disguise. The whole thing revolved around one ads experiment scenario and they squeezed five sub-questions out of it, covering hypothesis testing, effect sizes, power analysis, the works.

Questions Asked (5)

Q1

For an ads experiment comparing watch time per impression between two groups, how would you formally state the null and alternative hypotheses, and should you use a one-tailed or two-tailed test if the goal is to detect an increase?

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

I went with one-tailed since the product team only cares about detecting an increase, not a decrease.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clearly defining the metric (watch time per impression) and the two groups (control and treatment). Formally state the null hypothesis as no difference in means and the alternative as a difference, then discuss whether to use a one-tailed or two-tailed test based on the goal of detecting an increase. Conclude with a recommendation and justification.

Pro tip: Mention that while a one-tailed test is appropriate for detecting an increase, it requires strong justification and pre-registration to avoid p-hacking concerns; in practice, many companies default to two-tailed tests for robustness.

1. Define the metric and groups

Clarify that the metric is watch time per impression, and identify the control (A) and treatment (B) groups. Ensure the metric is continuous and the comparison is between means.

2. State the hypotheses

Null hypothesis (H0): μ_A = μ_B (no difference in mean watch time per impression). Alternative hypothesis (H1): μ_A ≠ μ_B for a two-tailed test, or μ_B > μ_A for a one-tailed test if detecting an increase.

3. Choose one-tailed vs two-tailed

If the goal is specifically to detect an increase (i.e., treatment improves watch time), a one-tailed test is appropriate. However, consider the trade-offs: one-tailed tests have more power to detect an increase but cannot detect a decrease, and may be seen as less rigorous.

4. Justify the choice

Explain that the choice depends on the research question and business context. If the business only cares about detecting an increase and a decrease would not lead to action, a one-tailed test is justified. Otherwise, use a two-tailed test.

5. Conclude with recommendation

Recommend a one-tailed test if the goal is strictly to detect an increase and it is pre-registered; otherwise, default to a two-tailed test for flexibility and to avoid criticism.

Key Points to Mention

  • Clear definition of the metric: watch time per impression
  • Formal hypothesis statements with proper notation (μ_A, μ_B)
  • Difference between one-tailed and two-tailed tests
  • Implications of choosing one-tailed vs two-tailed (power, directionality, rigor)
  • Business context: whether detecting a decrease is relevant
  • Pre-registration and avoiding p-hacking concerns

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

Q2

Compute the Welch t-statistic, approximate degrees of freedom, p-value, and 95% confidence interval for the difference in group means. Show your formulas and round to 3 decimal places.

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

This took me longer than I'd like to admit.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clearly stating the null and alternative hypotheses and the assumptions (independent groups, approximately normal or large samples). Then compute the group means, variances, and sample sizes, plug them into the Welch t-statistic formula, calculate the Welch–Satterthwaite degrees of freedom, find the p-value using the t-distribution, and construct the 95% confidence interval for the difference in means using the t-critical value. Show all formulas and round final answers to 3 decimal places.

Pro tip: Always interpret the results in the context of the business problem—e.g., whether the observed difference is practically significant—and mention that Welch's test is robust to unequal variances and sample sizes, which is common in A/B testing.

1. State hypotheses and assumptions

Define H0: μ1 = μ2 vs. H1: μ1 ≠ μ2 (or one-sided if appropriate). Verify assumptions: independent samples, approximately normal or large n, and note that Welch's test does not assume equal variances.

2. Compute sample statistics

Calculate the sample means (x̄1, x̄2), sample variances (s1², s2²), and sample sizes (n1, n2) for each group.

3. Calculate Welch t-statistic and degrees of freedom

Use the formula t = (x̄1 - x̄2) / sqrt(s1²/n1 + s2²/n2). Compute the Welch–Satterthwaite degrees of freedom: df = (s1²/n1 + s2²/n2)² / [ (s1²/n1)²/(n1-1) + (s2²/n2)²/(n2-1) ].

4. Find p-value and critical value

Using the t-distribution with the calculated df, find the two-sided p-value and the critical t-value for a 95% confidence level (t_{0.975, df}).

5. Construct confidence interval and interpret

Compute the 95% CI: (x̄1 - x̄2) ± t_{0.975, df} * sqrt(s1²/n1 + s2²/n2). Interpret the interval and p-value in the context of the experiment.

Key Points to Mention

  • Welch's t-test does not assume equal variances, making it more robust than Student's t-test for unequal sample sizes or variances.
  • The Welch–Satterthwaite equation for degrees of freedom approximates the t-distribution and is typically not an integer.
  • The standard error of the difference in means is sqrt(s1²/n1 + s2²/n2).
  • The p-value is obtained from the t-distribution with the calculated df; for a two-sided test, it's 2 * P(T > |t|).
  • The 95% confidence interval provides a range of plausible values for the true difference in means; if it excludes 0, the difference is statistically significant at α=0.05.
  • Always check assumptions and consider practical significance, not just statistical significance.

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

Q3

Calculate Cohen's d using the pooled standard deviation and interpret the effect size in practical terms.

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

Easy part of the whole problem.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clearly stating the formula for Cohen's d using pooled standard deviation, then walk through a concrete example with numbers to show the calculation. Finally, interpret the resulting effect size in practical terms, relating it to the business context such as the magnitude of improvement in a metric.

Pro tip: Always relate the effect size back to business impact—e.g., a small effect might still be valuable if it translates to significant revenue. Also, mention that Cohen's d assumes normality and equal variances, so check those assumptions.

1. Define Cohen's d and its purpose

Explain that Cohen's d measures the standardized difference between two means, useful for comparing effect sizes across different scales. Mention that it's calculated as the difference in means divided by the pooled standard deviation.

2. Present the formula for pooled standard deviation

State the formula: s_p = sqrt(((n1-1)s1^2 + (n2-1)s2^2) / (n1+n2-2)), where n1, n2 are sample sizes and s1, s2 are standard deviations. Then Cohen's d = (mean1 - mean2) / s_p.

3. Walk through a numerical example

Use a simple example, such as two groups with means 10 and 8, standard deviations 2 and 2.5, and sample sizes 30 each. Compute pooled SD and then d, showing each step.

4. Interpret the effect size

Refer to Cohen's benchmarks: 0.2 small, 0.5 medium, 0.8 large. Explain what the calculated d means in context—e.g., a d of 0.8 indicates a large effect, meaning the difference is substantial relative to variability.

5. Relate to practical business impact

Discuss how this effect size translates to real-world outcomes, such as a lift in conversion rate or revenue. Emphasize that statistical significance alone isn't enough; effect size shows practical significance.

Key Points to Mention

  • Formula for Cohen's d and pooled standard deviation
  • Assumptions: normality, homogeneity of variance
  • Cohen's benchmarks for small, medium, large effects
  • Difference between statistical significance and practical significance
  • How to compute and interpret in the context of A/B testing
  • Potential limitations: sensitivity to sample size, non-normal data

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

Q4

If you had pre-registered a one-tailed test for an increase, how does the p-value change, and what are the risks of mis-specifying the direction beforehand?

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

The p-value halves when you go one-tailed, so if two-tailed was around 0.023 the one-tailed version is about 0.012.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by explaining that a one-tailed test concentrates the entire alpha in one direction, so for an observed effect in the pre-registered direction, the p-value is exactly half of the two-tailed p-value. Then discuss the risks: if the true effect is in the opposite direction, the one-tailed test will never detect it, and if you choose the direction after seeing the data, you inflate Type I error. Finally, emphasize the importance of pre-registration and the trade-offs in practical A/B testing.

Pro tip: Mention that in industry, stakeholders often want to know if a change is 'better' or 'worse', so a one-tailed test can be tempting, but it's safer to use a two-tailed test and report directional conclusions only if the effect is significant in the expected direction. This shows you balance statistical rigor with business needs.

1. Define one-tailed vs two-tailed tests

Clarify that a one-tailed test allocates all alpha to one direction, while a two-tailed test splits alpha across both tails. This sets the foundation for understanding p-value changes.

2. Explain the p-value change

State that for an effect in the pre-registered direction, the one-tailed p-value is half the two-tailed p-value. For an effect in the opposite direction, the one-tailed p-value is 1 minus half the two-tailed p-value (i.e., non-significant).

3. Discuss risks of mis-specifying direction

Highlight that if the true effect is in the opposite direction, you lose power to detect it and may incorrectly conclude no effect. Also, if you choose the direction after peeking at data, you inflate Type I error and invalidate the test.

4. Emphasize pre-registration and practical implications

Stress that pre-registration is crucial to avoid p-hacking. In A/B testing, consider whether the business question truly warrants a directional hypothesis or if a two-tailed test is more robust.

5. Conclude with a balanced recommendation

Recommend using one-tailed tests only when there is strong theoretical or practical justification for a single direction, and always pre-register. Otherwise, default to two-tailed tests for flexibility and integrity.

Key Points to Mention

  • One-tailed p-value is half of two-tailed p-value when the effect is in the pre-registered direction.
  • If the effect is in the opposite direction, the one-tailed test will not detect it, leading to a false negative.
  • Choosing the direction after seeing the data inflates Type I error rate and is a form of p-hacking.
  • Pre-registration of the direction is essential to maintain statistical validity.
  • One-tailed tests can be appropriate when there is a strong prior or business reason to expect only one direction.
  • In A/B testing, two-tailed tests are generally safer and more informative for stakeholders.

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

Q5

Using alpha of 0.05 two-sided and 80% power, what per-group sample size do you need to detect a 2-second difference assuming a common standard deviation of about 30.5? Walk through the formula.

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

Power analysis questions always make me nervous because I mix up the z-score values.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by stating the two-sample t-test sample size formula for comparing means, then plug in the given values (alpha=0.05 two-sided, power=80%, delta=2, sigma=30.5). Compute the required per-group sample size and round up to the nearest integer, briefly explaining each component.

Pro tip: Mention that this calculation assumes a two-sided test and equal variances; in practice, you might adjust for unequal variances or use a z-test approximation for large samples. Also, note that the formula gives the minimum sample size, so rounding up is necessary.

1. Identify the formula

Use the standard formula for sample size per group in a two-sample t-test: n = 2 * (Z_{1-α/2} + Z_{1-β})^2 * σ^2 / Δ^2, where Z values are from the standard normal distribution.

2. Determine critical values

For α=0.05 two-sided, Z_{1-α/2} = 1.96. For 80% power, Z_{1-β} = 0.84. Sum them: 1.96 + 0.84 = 2.80.

3. Plug in values and compute

Substitute σ=30.5, Δ=2, and the sum of Z's into the formula: n = 2 * (2.80)^2 * (30.5)^2 / (2)^2. Calculate step by step: (2.80)^2 = 7.84, (30.5)^2 = 930.25, so numerator = 2 * 7.84 * 930.25 = 14586.32; denominator = 4; n = 3646.58.

4. Round up and interpret

Since sample size must be an integer, round up to 3647 per group. Explain that this is the minimum required to achieve 80% power at the given significance level.

Key Points to Mention

  • Two-sided test and 80% power correspond to Z-values 1.96 and 0.84 respectively.
  • The formula assumes equal variances and normally distributed data; for large samples, the t-distribution approximates the normal.
  • The effect size (Δ/σ) is 2/30.5 ≈ 0.0656, which is small, hence the large sample size.
  • Sample size is per group; total sample size would be double (7294).
  • Rounding up is necessary because you cannot have a fraction of a participant.
  • In practice, consider dropout or non-compliance by inflating the sample size.

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