What it means

A hypothesis is a clear statement about how the world works that can be checked by observation or experiment.
Testable means you can collect data that bear on the claim.
Falsifiable means there is a possible result that would show the claim is wrong.

If no result could ever count against a claim, it is not a scientific hypothesis. It may be a belief, a definition, or a vague idea, but it is not science.

How to apply it

1. State the claim precisely
   Use measurable variables, a specific context, and a predicted pattern.
   Example: “Drinking 500 ml of coffee within 30 minutes of waking increases reaction time on a 2 minute tapping test by at least 5 percent compared with water.”
2. Identify what would disconfirm the claim
   Example: “If the coffee group’s average improvement is under 5 percent or worse than water, the hypothesis is not supported.”
3. Plan methods before seeing results
   Choose sample size, measures, and analysis in advance to avoid moving the target.
4. Collect data that could go either way
   Use controls, randomization where possible, and record all outcomes, not only the pretty ones.
5. Compare prediction to outcome
   If results match, you have support, not proof. If results contradict, you have learned something, which is progress.
6. Revise and retest
   Update the hypothesis, tighten methods, and run another check. Iteration is the engine.

Why it works

• It forces clarity
  A falsifiable hypothesis removes wiggle room and makes disagreements solvable with evidence.
• It resists bias
  Predefining what would count against you reduces selective memory and cherry picking.
• It compounds knowledge
  Disconfirmed ideas are retired or refined, which keeps the knowledge base clean and usable.
• It invites replication
  Clear tests can be repeated by others, building trust and catching errors.

Everyday examples

• Fitness
  Hypothesis: “Three 20 minute brisk walks per week will lower my resting heart rate by 5 beats per minute within 6 weeks.”
  Test: Track resting heart rate every morning, compare week 1 to week 6.
  Falsifier: Reduction under 5 beats or an increase.
• Sleep
  Hypothesis: “Stopping screens at 9 pm improves my sleep efficiency to at least 90 percent within 10 days.”
  Test: Use a sleep app or watch.
  Falsifier: Efficiency stays below 90 percent.
• Budgeting
  Hypothesis: “Writing every purchase before paying reduces monthly discretionary spending by 15 percent.”
  Test: Compare last month to this month with the rule.
  Falsifier: Reduction under 15 percent.
• Learning
  Hypothesis: “Daily 15 minute spaced flashcards will raise quiz scores by at least 10 percent after two weeks.”
  Test: Baseline quiz today, repeat after two weeks.
  Falsifier: Improvement under 10 percent.
• Productivity
  Hypothesis: “Setting a 25 minute timer and closing messaging apps will raise deep work time to 90 minutes per day.”
  Test: Track deep work minutes.
  Falsifier: Average stays below 90.

Common pitfalls to avoid

• Vague wording
  Words like “better,” “often,” or “more” without numbers block testing.
• Moving goalposts
  Changing targets after seeing results removes falsifiability.
• Unmeasured confounders
  If something else could explain the result, refine the design or add controls.
• One off victories
  A single success can be luck. Seek stable effects across time.

Quick checklist

• Is the claim specific, measurable, and time bound
• Do I know exactly what result would count against it
• Did I lock methods before collecting data
• Can someone else repeat this test from my notes
• Will I act on the result, even if it goes against my hope

Closing thought

Science advances by making bold, precise bets that reality can settle. When your hypotheses are testable and falsifiable, every outcome teaches you something useful.