Calculate conditional probability for user retention

Last updated: April 12, 2026

Quick Overview

Given the following scenario about user conversion, calculate the the p-value.

Mastercard
Analytics & Experimentation
Data Scientist
Mastercard
April 12, 2026
Data Scientist
Onsite
Analytics & Experimentation
Medium

1

4

98 solved


Given the following scenario about user conversion, calculate the the p-value.

This analytics question from Mastercard's Onsite tests your ability to think critically about data. The interviewer expects you to consider confounding variables, selection bias, and the difference between correlation and causation.

What the Interviewer Expects
  • Design a rigorous experiment with proper randomization and sample size calculation
  • Define primary and guardrail metrics with clear rationale
  • Address novelty effects, network effects, and interference
  • Segment results appropriately and identify heterogeneous treatment effects
  • Propose follow-up analyses when results are ambiguous
Key Topics to Cover
Sample size and power calculation
Long-term vs short-term metrics
Segmentation and heterogeneous effects
Funnel analysis and cohort analysis
Network effects and interference
Guardrail metrics
How to Approach This
  1. Define success metrics carefully. A good metric is measurable, actionable, and aligned with business goals.
  2. Run experiments long enough to account for novelty effects and weekly seasonality.
  3. Use funnel analysis to identify where users drop off for maximum optimization impact.
  4. Segment results by key dimensions (platform, country, user cohort) to catch hidden patterns.
  5. Consider network effects and interference between treatment and control groups.
Possible Follow-up Questions
  • How would you handle an experiment where the control and treatment groups are different sizes?
  • How would you handle seasonality in your experiment?
  • What if you discover a bug in the logging during the experiment?
  • What would you do if a stakeholder wants to end the experiment early because initial results look good?
Sharpen Your Skills on Codemia

Practice similar problems with our interactive workspace, get AI feedback, and track your progress.

Browse Analytics Questions
Sample Answer
Defining the Analysis Framework

Start by framing the problem clearly: 1. **What question are we answering?** Be specific about the hypothesis or business question. 2. **What data d...

Analysis Approach

**Exploratory Analysis**: Start with aggregate trends, then segment by key dimensions (platform, country, user cohort, channel). Look for heterogeneou...


Submit Your Answer
Markdown supported

Related Questions