Apps

🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!

Regression Discontinuity Misuse

Also Known As: RDD Misspecification
Discourse Mechanics ID: regression_discontinuity_misuse

Definition

Errors arising from improper application of regression discontinuity designs, including incorrect functional form assumptions, inappropriate bandwidth selection, or failure to detect manipulation of the assignment variable near the cutoff. These errors can produce spurious treatment effects.

Examples

A study of a scholarship program uses a test score cutoff but fits a linear model to data that is actually curved, producing a false 'jump' at the cutoff that is really a modeling artifact.

A city evaluates a speed camera program by comparing accident rates just above and below the 30 mph speed limit used to determine camera placement. Researchers use a very narrow bandwidth of only a few observations near the cutoff, producing an unstable estimate that swings wildly with the inclusion or exclusion of a single data point, and falsely conclude the cameras have no effect.

A study assessing whether students who barely pass a remedial math course (cutoff: 60%) earn higher wages later in life fits a single straight regression line across a dataset where the true relationship curves sharply near the cutoff. The apparent 'jump' in wages at the threshold is largely an artifact of forcing a linear model onto nonlinear data, not evidence that the course itself raises earnings.

Verification Steps
Verification Steps
Binary yes/no questions that an AI must answer to detect a reasoning pattern in a text.
Each of the 452 aspects has verification steps — simple yes/no questions designed to systematically detect whether a pattern appears in a text. For ad hominem: "Does the argument attack a person rather than their claim?" For false dichotomy: "Are only two options presented when more exist?" This ensures consistent, reproducible analysis.

Binary (yes/no) questions an LLM must answer to identify this aspect:

  1. 1

    Is a treatment or intervention assigned based on a cutoff value of a continuous variable?

    Type: binary
  2. 2

    Is the effect estimated by comparing groups just above and below the cutoff?

    Type: binary
  3. 3

    Are assumptions about the functional form near the cutoff (linearity, bandwidth) being validated?

    Type: binary
  4. 4

    Could manipulation of the running variable near the cutoff invalidate the design?

    Type: binary
Deep Dive
The expandable detail section on each aspect page with examples, psychology, and counter-strategies.
The Deep Dive section provides in-depth information about each aspect: a real-world example showing the pattern in action, an explanation of why it works psychologically, practical advice on how to counter it, alternative names, and links to related aspects.

Hierarchical Context