M1-056M1: Mathematics & Statistics FundamentalsCausal InferenceEasy
⚡Flashcard Drill➔Mastery:

Define Causal Effects and the Potential Outcomes Framework (Rubin Causal Model)

🎯 Active Recall Self-Check (Draft First, Verify After)
Login-free practice tracker: recall core answer, audit pitfalls, and log review status.
Step 1 · Draft your concise answer & assumptions
Step 2 · Check Pitfalls & Follow-up Inquiries
Next: M1-057 →
📐 Mathematical Definition
τi=Yi(1)−Yi(0),ATE=E[Y(1)−Y(0)]\tau_i=Y_i(1)-Y_i(0),\qquad \mathrm{ATE}=\mathbb E[Y(1)-Y(0)]
⚡ One-Sentence Core Answer
The Rubin Causal Model defines causal effect as the difference between potential outcomes under treatment and control: Yi(1)−Yi(0)Y_i(1) - Y_i(0); the Fundamental Problem of Causal Inference is that we can only ever observe one potential outcome for any individual.

📌 Key Takeaways

  • •
    Potential Outcomes: Yi(1)Y_i(1) (outcome if treated) and Yi(0)Y_i(0) (outcome if control).
  • •
    Observed outcome: Yi=TiYi(1)+(1−Ti)Yi(0)Y_i = T_i Y_i(1) + (1 - T_i) Y_i(0), where Ti∈{0,1}T_i \in \{0, 1\}.
  • •
    Average Treatment Effect (ATE): τATE=E[Y(1)−Y(0)]\tau_{\text{ATE}} = E[Y(1) - Y(0)].
  • •
    SUTVA (Stable Unit Treatment Value Assumption): No interference between units and no hidden variations of treatment.

📐 Mathematical Derivations

Observed difference in group means decomposes into treatment effect plus selection bias: E[Y∣T=1]−E[Y∣T=0]=E[Y(1)∣T=1]−E[Y(0)∣T=0]=E[Y(1)∣T=1]−E[Y(0)∣T=1]+E[Y(0)∣T=1]−E[Y(0)∣T=0]=τATT+Selection BiasE[Y\mid T=1] - E[Y\mid T=0] = E[Y(1)\mid T=1] - E[Y(0)\mid T=0] = E[Y(1)\mid T=1] - E[Y(0)\mid T=1] + E[Y(0)\mid T=1] - E[Y(0)\mid T=0] = \tau_{\text{ATT}} + \text{Selection Bias}, where Selection Bias=E[Y(0)∣T=1]−E[Y(0)∣T=0]\text{Selection Bias} = E[Y(0)\mid T=1] - E[Y(0)\mid T=0]. In observational data, treated units typically have higher baseline outcomes even without treatment (e.g. Users who adopt a premium feature have higher organic activity), so selection bias ≠0\ne 0. Randomized Controlled Trials (RCT) enforce T⊥(Y(0),Y(1))T \perp (Y(0), Y(1)), eliminating selection bias so E[Y(0)∣T=1]=E[Y(0)∣T=0]=E[Y(0)]E[Y(0)\mid T=1] = E[Y(0)\mid T=0] = E[Y(0)], guaranteeing that observed difference equals true ATE.

🏭 Production Trade-offs

SUTVA requirements: (1) No Interference: User ii's outcome is unaffected by whether user jj receives treatment. In ride-sharing (Uber/Lyft) and social networks, treating drivers or users spills over to competitors, violating SUTVA and requiring switchback or cluster randomization. (2) Consistency: The treatment must be a single, uniform intervention without ambiguous variations.
⚠️ Common Interview Pitfalls
  • ✕
    Confusing association E[Y∣T=1]−E[Y∣T=0]E[Y\mid T=1] - E[Y\mid T=0] with causation E[Y(1)−Y(0)]E[Y(1) - Y(0)] in observational datasets.
  • ✕
    Ignoring network spillover interference in two-sided marketplaces, violating SUTVA.
🎯 Interviewer Follow-up Inquiries
  • ?
    What is the difference between Average Treatment Effect (ATE), Average Treatment Effect on the Treated (ATT), and Conditional Average Treatment Effect (CATE)?
  • ?
    How do spillover effects violate SUTVA in marketplace experiments?
⚡
Test muscle memory on this topic via 3-Layer Flashcards
Active Recall: Mentally formulate Prompt ➔ Math Derivation ➔ Benchmark Implementation in 3 minutes.
🎯 Drill This Card Now➔
📚

Associated Knowledge Base Guides & Mindmaps

← PreviousM1-055: Analyze the Statistical Diagnostics When CTR Increases but CVR Remains Flat in an Online Experiment📋 Back to BankNext →M1-057: Explain Confounders, Mediators, and Colliders in Causal DAGs and Which Must or Must Not Be Controlled