Your question is Experiment Pitfalls: Network Effects and SRM. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
Walk me through common experimentation pitfalls such as network effects, spillover, and sample ratio mismatch (SRM), and explain how you would detect and mitigate them.
Your answer should distinguish interference from assignment or instrumentation failures, describe appropriate diagnostic tests, and explain how the unit of randomization may need to change. Cover what you would do before launch, during monitoring, and before trusting the final result. Include how you would handle an experiment that shows SRM or evidence of spillover.
