Prepia
All posts

Behavioral Health · 8 min read

Test anxiety: what the evidence supports, and what it doesn’t

Breathing apps, lucky pens, last-minute cramming — we sorted the interventions with real effect sizes from the comforting rituals.

Search “test anxiety” and you'll find breathing apps, manifestation scripts, lucky pens, and someone's uncle insisting you eat two bananas. Some of it works. Most of it is comfort theater. The difference matters when you have three weeks left and finite evenings.

We went through the intervention literature and sorted it into what has real effect sizes, what doesn't, and what to actually do in the time you have.

What the evidence supports

  • Practice under realistic conditions. Full-length, timed, same time of day as your real exam. Exposure is the closest thing anxiety research has to a sure bet — the exam room loses its power when it isn't novel.
  • Expressive writing. Ten minutes before a high-stakes test, writing out your worries measurably improves performance for anxious test-takers — the effect replicated across classroom studies. It works by unloading the worry loop that competes with working memory.
  • Reappraisal. Telling yourself “I'm excited” outperforms “calm down” in controlled studies. Arousal isn't the enemy; the interpretation is. A racing heart before an exam is also what readiness feels like.
  • Slow breathing, briefly. A few minutes of paced breathing (around six breaths a minute) reliably drops acute physiological arousal. It won't cure anything — it buys you a usable ten minutes.
  • Sleep, boringly. Sleep regularity in the final week predicts performance better than total study hours in the final week. This one has the least glamour and the most data.

What doesn't hold up

  • Last-minute cramming as anxiety management. It feels like control; it measurably degrades both sleep and next-day recall. The anxiety is telling you to do it precisely because it's the familiar move.
  • Avoiding practice tests “to protect confidence.” Avoidance is the maintenance mechanism of anxiety, not a treatment for it. Every skipped simulation makes the real one more alien.
  • Starting an eight-week mindfulness program three days out. The research on mindfulness for anxiety is real — over months. As an exam-week intervention it mostly adds a task you're now also behind on.
  • Rituals — the pen, the playlist, the seat. Harmless, occasionally soothing, not treatment. Keep the pen; just don't confuse it for a plan.

A two-week protocol you can actually run

Two weeks out: schedule three full-length simulations at your real exam's start time, spaced three or four days apart. Keep your wake time fixed the whole stretch. The night before each sim — and the real thing — write your worries down for ten minutes, then stop. Morning of, five minutes of slow breathing in the car or the lobby, and walk in slightly early, not absurdly early; forty minutes of pacing the hallway undoes the breathing.