← CVS Health Interview Insights

CVS Health·Data Scientist·Technical Phone Screen·Intermediate

Intermediate
May 2026

Summary

CVS Health data scientist interview that leaned heavily on applied stats. Three questions, all quantitative, no fluff about your background or why you want the job. If your stats fundamentals are rusty this will hurt.

Questions Asked (3)

Q1

You have a sample of 64 observations from a normal population with unknown variance, a sample mean of 102, and a sample standard deviation of 16. Construct a 95% confidence interval for the population mean. State the formula you use and the degrees of freedom for your t critical value.

Product Analytics & Metrics
Author's notes

Pretty standard but the unknown variance part is the key detail.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

First, identify that the population variance is unknown and the sample size is 64, so use the t-distribution with 63 degrees of freedom. Then, state the confidence interval formula: x̄ ± t_(α/2, df) * (s/√n), and plug in the values to compute the interval. Finally, interpret the interval in the context of the problem.

Pro tip: Mention that with n=64, the t and z critical values are nearly identical (t≈2.00 vs z=1.96), but using t is technically correct and shows statistical rigor. Also, briefly interpret the interval in plain language for a non-technical audience, which is valuable in a healthcare company like CVS Health.

1. Identify the appropriate distribution

Since the population variance is unknown and the sample size is 64, use the t-distribution with n-1 = 63 degrees of freedom. Note that the sample size is large enough that the t-distribution approximates the normal, but the t is still the correct choice.

2. State the confidence interval formula

The formula for a confidence interval for the population mean with unknown variance is: x̄ ± t_(α/2, df) * (s/√n), where x̄ is the sample mean, s is the sample standard deviation, n is the sample size, and t_(α/2, df) is the critical value from the t-distribution.

3. Find the critical value

For a 95% confidence level, α = 0.05, so α/2 = 0.025. With 63 degrees of freedom, the critical t-value is approximately 1.998 (or 2.00). You can mention that you would look this up in a t-table or compute it using software.

4. Compute the margin of error and interval

Calculate the standard error: s/√n = 16/√64 = 16/8 = 2. Then margin of error = t * SE ≈ 1.998 * 2 = 3.996. The confidence interval is 102 ± 3.996, which gives (98.004, 105.996).

5. Interpret the interval

Interpret the interval in context: We are 95% confident that the true population mean lies between approximately 98.0 and 106.0. This means if we repeated this sampling process many times, 95% of the intervals constructed would contain the true mean.

Key Points to Mention

  • Use of t-distribution because population variance is unknown.
  • Degrees of freedom = n - 1 = 63.
  • Formula: x̄ ± t_(α/2, df) * (s/√n).
  • Calculation of standard error: s/√n = 16/8 = 2.
  • Critical t-value for 95% confidence and 63 df is approximately 2.00.
  • Interpretation of the confidence interval in the context of the problem.

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

Q2

You measure two variables on 100 observations and get a Pearson correlation of 0.25. Test whether the true population correlation is zero at the 0.05 significance level. Show your test statistic, degrees of freedom, the two-sided p-value, and your conclusion.

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

This one tripped me up more than it should have.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Use the t-test for Pearson correlation: compute t = r * sqrt((n-2)/(1-r^2)) with df = n-2, then find the two-sided p-value from the t-distribution. Compare p to 0.05 and state the conclusion in context.

Pro tip: Mention that with n=100, the test has reasonable power to detect moderate correlations, but always check assumptions like linearity and normality. Also, note that a non-significant result doesn't prove the null; it just means insufficient evidence.

1. State hypotheses and significance level

Define H0: ρ = 0 vs. H1: ρ ≠ 0, and set α = 0.05.

2. Compute the test statistic

Calculate t = r * sqrt((n-2)/(1-r^2)) using r = 0.25 and n = 100.

3. Determine degrees of freedom and p-value

df = n - 2 = 98. Find the two-sided p-value from the t-distribution with 98 df.

4. Compare p-value to α and conclude

If p < 0.05, reject H0; otherwise fail to reject. Interpret in context of the variables.

Key Points to Mention

  • Formula for the t-test for correlation: t = r * sqrt((n-2)/(1-r^2))
  • Degrees of freedom = n - 2 = 98
  • Two-sided p-value from t-distribution with 98 df
  • Conclusion: fail to reject H0 if p ≥ 0.05, meaning no significant evidence of non-zero correlation
  • Assumptions: linear relationship, bivariate normality, no outliers
  • Effect size and practical significance: r = 0.25 is a weak correlation

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

Q3

A process follows a normal distribution with mean 50 and standard deviation 5. Find the value t such that only the top 2.5% of items exceed it, and explain what that means in context.

Product Analytics & Metrics
Author's notes

Easiest of the three.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

First, recognize that the top 2.5% corresponds to a z-score of approximately 1.96. Then, use the formula t = μ + zσ to compute the threshold value. Finally, interpret the result in the context of the process, explaining that only 2.5% of items are expected to exceed this value.

Pro tip: When explaining the context, relate it to business implications, such as identifying high-cost patients or outliers, to demonstrate practical understanding. Also, mention that the z-score for 2.5% is a common critical value in statistics, often used for 95% confidence intervals.

1. Identify the given parameters

State the mean (μ = 50) and standard deviation (σ = 5) of the normal distribution.

2. Determine the z-score for the top 2.5%

Recall that the top 2.5% corresponds to a z-score of approximately 1.96 (since 97.5% of the distribution lies below it).

3. Calculate the threshold value t

Use the formula t = μ + zσ = 50 + 1.96 * 5 = 59.8.

4. Interpret the result in context

Explain that only 2.5% of items from this process are expected to have values greater than 59.8. In a business context, this could represent the top 2.5% of patients by some metric, helping to identify outliers or high-risk groups.

Key Points to Mention

  • Normal distribution properties
  • Z-score calculation and interpretation
  • The 68-95-99.7 rule (empirical rule)
  • The specific z-score for 2.5% (1.96)
  • Formula for converting z-score to raw value: X = μ + zσ
  • Business context: identifying top 2.5% as outliers or high-priority cases

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