← Apple Interview Insights

Apple·Data Scientist·Technical Phone Screen·Intermediate

Intermediate
May 2026

Summary

Apple data scientist interview with a statistics question that felt deceptively textbook until I was actually in it. The focus was on distribution theory and when approximations hold, which sounds like review material until someone's staring at you waiting for a derivation.

Questions Asked (1)

Q1

What are the key differences between the Normal and Poisson distributions, and when can you approximate a Poisson with a Normal? Derive the mean and variance for both.

Data ModelingTechnical Trade-offs
Author's notes

I started okay on the conceptual stuff.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clearly defining both distributions, emphasizing their support and parameters, then derive the mean and variance for each using first principles. Finally, explain the conditions under which a Poisson can be approximated by a Normal, including the role of λ and continuity correction.

Pro tip: Mention that the Poisson is a discrete distribution often used for count data, while the Normal is continuous and arises from the Central Limit Theorem; this shows you understand when each is appropriate in practice.

1. Define the distributions

State that the Normal is a continuous distribution defined on (-∞, ∞) with parameters μ and σ², while the Poisson is a discrete distribution on non-negative integers with parameter λ.

2. Derive mean and variance for Normal

For Normal, the mean is μ and variance is σ², which can be derived from the probability density function or by using the moment-generating function.

3. Derive mean and variance for Poisson

For Poisson, both mean and variance equal λ. Derive using the probability mass function and the definition of expectation, or via the moment-generating function.

4. Explain approximation conditions

The Poisson can be approximated by a Normal when λ is large (typically λ > 20 or 30). Use continuity correction for better accuracy, and note that the Normal approximation improves as λ increases due to the Central Limit Theorem.

5. Summarize key differences

Highlight that Normal is continuous and symmetric, while Poisson is discrete and skewed for small λ. Also note that Poisson models counts per unit time/space, while Normal models continuous measurements.

Key Points to Mention

  • Normal distribution is continuous, symmetric, and defined by mean μ and variance σ².
  • Poisson distribution is discrete, models counts, and has mean = variance = λ.
  • Derivation of Poisson mean and variance using the probability mass function or MGF.
  • Approximation condition: λ > 20 (or 30) for Normal approximation.
  • Use of continuity correction when approximating discrete with continuous.
  • Central Limit Theorem justifies the approximation as λ grows.

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