This question evaluates a data scientist's experiment design competency, including A/B testing fundamentals, randomization unit choice, stratification/blocking, sample size and power considerations, guardrail metrics, and exposure/interference handling for online recommender systems.

A social app shows hashtag recommendations to users while composing posts. A new algorithm is proposed to increase hashtag adoption and downstream engagement without harming reliability or content quality.
Design and run an online experiment to evaluate the new recommender. Focus on how you would choose test and control groups, and address:
Login required