Skip to content

Small experiment

pageplanreflectconverge

Test a feared prediction with a small, safe action in the real world, then compare what happened with what you expected.

Variants

  • default

Data fields

The fields a spec's page can set for this component (see switchback show small-experiment --json):

  • aversion

Grounding

Helps: when an aversion rests on a prediction you can test cheaply: writing the prediction down, running one small safe probe and comparing the result at the return follows how exposure work is structured, which helps when self-guided but less than with a guide; the comparison step's own contribution is untested

Backfires: when it is used on a diagnosed anxiety problem, which it does not treat, when the probe is too big to feel safe, or when the result is never set against the prediction; whether an unguided probe can make the fear worse is untested

What the grades mean