##### Scenario You are part of the analytics team at a social media platform. Recently, it has been observed that there is a positive correlation between the number of page visits to an advertisement and the reports labeling it as bad content. An investigation is warranted to understand this phenomenon better. ##### Question Define what constitutes a 'top bad advertiser' using SQL. What metrics and conditions would you apply to identify advertisers with high incidents of reported ads? Propose a method to investigate the positive correlation between ad page visits and the frequency of reports. Develop a hypothesis and outline the steps you would take to validate or refute it. ##### Hints Consider the appropriateness of correlation vs. causation. Analyze potential factors contributing to higher report rates, including ad content, engagement types, and user demographics. Describe the data fields and SQL queries you might use to break down and assess these patterns.

This question evaluates a data scientist's competency in operational analytics, including defining and aggregating metrics in SQL and applying causal reasoning to interpret correlations between ad exposures and user reports.

How do I approach Analytics & Experimentation interview questions?

Analytics & Experimentation questions require understanding of core concepts and practice. PracHub provides solutions with explanations to help you master analytics & experimentation interviews.

What difficulty level is this interview question?

This is a medium difficulty Analytics & Experimentation question, commonly asked during Technical Screen rounds at TikTok.

What role is this question designed for?

This question is commonly asked for Data Scientist candidates at TikTok during technical interviews.

Investigating Correlation Between Ad Visits and Reports

Scenario

You are on the analytics team at a large social media platform. A positive correlation has been observed between the number of page visits to an advertisement and the number of user reports labeling it as bad content. An investigation is needed to understand what drives this pattern and to define how to flag top bad advertisers.

Task 1: Define 'top bad advertiser' in SQL

Propose clear metrics and conditions to identify advertisers with high incidents of reported ads.
Include rate-based metrics (not just counts), minimum volume thresholds, deduplication rules, and any severity weighting.
Provide example SQL that aggregates events and ranks advertisers accordingly.

Task 2: Investigate the visit–report correlation

Propose a hypothesis for why ad page visits are positively correlated with reports.
Outline a plan to validate or refute the hypothesis, calling out correlation vs. causation.
Describe the data fields and breakdowns you would use (e.g., ad content, engagement types, user demographics, placement, time).
Include example SQL to compute and analyze these patterns.

Hints

Do not conflate correlation with causation; control for exposure and composition of traffic.
Consider confounders: ad category, creative type, placement, region, device, time, and user segments.
Think about unique reporters vs. total reports, verification by moderation, and report timing relative to visits.

Scenario

Task 1: Define 'top bad advertiser' in SQL

Propose clear metrics and conditions to identify advertisers with high incidents of reported ads.

Include rate-based metrics (not just counts), minimum volume thresholds, deduplication rules, and any severity weighting.

Provide example SQL that aggregates events and ranks advertisers accordingly.

Task 2: Investigate the visit–report correlation

Propose a hypothesis for why ad page visits are positively correlated with reports.

Outline a plan to validate or refute the hypothesis, calling out correlation vs. causation.

Describe the data fields and breakdowns you would use (e.g., ad content, engagement types, user demographics, placement, time).

Include example SQL to compute and analyze these patterns.

Hints

Do not conflate correlation with causation; control for exposure and composition of traffic.

Consider confounders: ad category, creative type, placement, region, device, time, and user segments.

Think about unique reporters vs. total reports, verification by moderation, and report timing relative to visits.

Investigating Correlation Between Ad Visits and Reports

Quick Overview

Scenario

Task 1: Define 'top bad advertiser' in SQL

Task 2: Investigate the visit–report correlation

Hints

Solution

Comments (0)

Investigating Correlation Between Ad Visits and Reports

Quick Overview

Scenario

Task 1: Define 'top bad advertiser' in SQL

Task 2: Investigate the visit–report correlation

Hints

Solution

Comments (0)