Regression discontinuity design

RDD uses a sharp threshold for treatment assignment

Regression discontinuity designs (RDD) rely on a sharp cutoff or threshold to assign treatments. Observations just above and below this threshold are compared to estimate the treatment effect. This method helps in situations where random assignment is not possible.

Example

In evaluating scholarship programs, students scoring just above or below a certain GPA threshold are assigned to receive scholarships or not, respectively. Researchers then compare outcomes of students near this cutoff to estimate the scholarship's impact.

Remember this

RDD provides a way to infer causal effects in quasi-experimental settings, which is crucial for fields like economics and psychology where randomization is often impractical.

Text adapted from Wikipedia, licensed under CC BY-SA 4.0.

Related concepts

to standardize: when you need zero mean and unit variance for gradient-based optimization

Why do we need to make data uniform before training a model?

score matching does: learns the gradient of the log-density without normalizing

Propensity score matching reduces bias in treatment effect estimates

denoising score matching does: learns to denoise, which equals learning the score

Propensity score matching (PSM) reduces bias in treatment effect estimates

Adam has bias correction: divides by (1-β^t) in early steps

Why do we sometimes need to fix mistakes in computer decisions?

LoRA (machine learning)

LoRA uses r << d for efficient adaptation

Langevin dynamics does: adds noise to gradient descent to sample from a distribution

Langevin dynamics uses stochastic differential equations

Swipe through 100 ML concepts daily

Open Pocket Polymath