1. When Does a Discovery Reflect Real Signal, Not Dependency-Inflated Noise?

    Large-scale testing where the textbook null fails once dependence enters the data.

  2. What Does a Predictive-Performance Number Mean Once Feature Selection Is Part of the Pipeline?

    Validation that counts the selection step instead of treating it as free.

  3. What Has to Be True Before an Institution Acts on a Causal Claim?

    Population, design and evidence strong enough for a regulator, an agency or a hospital.

  4. Can a Prediction of Disease Risk Carry Uncertainty a Clinician Can Act On?

    Multiclass conformal prediction for disease development and progression.

    New work, not yet built.

  5. Can Disease Progression Be Stratified Without Forcing It Into Fixed Cutoffs?

    Clustered survival where some patients never reach the event at all.

  6. Can Biological Age Close the Gap Between Lifespan and Healthspan?

    The throughline: whether a corrected biological-age gap predicts disease on its own.

    Under way as replications, not papers yet.