Preregistration is the practice of recording key research decisions before the results are known. The core aim of Science mastery is not to make research inflexible. It is to help students understand the difference between a hypothesis tested as planned and an interesting pattern discovered after looking at the data.
For students and parents searching for preregistration, preregistration in Science, registered research, analysis plan, p-hacking, research transparency or registered reports, the most useful principle is this: preregistration creates a timestamped record of what the researchers intended to test before knowing which results would look exciting.
That makes hindsight easier to detect and scientific claims easier to interpret.
The 60-Second Preregistration
A preregistration may specify:
- research question;
- hypotheses;
- primary outcomes;
- sample size;
- exclusion criteria;
- variables;
- statistical tests;
- stopping rules.
The details depend on the study.
Wait, What? Preregistration Does Not Ban Exploration?
Correct.
Exploration is a vital part of Science.
Preregistration simply distinguishes:
Confirmatory analysis: planned before seeing the results.
Exploratory analysis: discovered or decided after inspecting the data.
Both can be valuable.
The problem arises when exploratory findings are presented as though they were predicted in advance.
Why Timing Matters
Once researchers see the data, many choices can be influenced by the result.
Examples:
- which outcome to emphasise;
- which participants to exclude;
- which subgroup to analyse;
- which statistical model to use;
- whether to stop collecting data.
Preregistration records these decisions before the temptation of hindsight appears.
Preregistration and Hypotheses
A preregistered hypothesis should be:
- specific;
- measurable;
- linked to defined variables;
- paired with an analysis plan.
See Null Hypothesis and Alternative Hypothesis.
Primary and Secondary Outcomes
A study may measure many things.
Preregistration can identify:
- primary outcome: the main result the study is designed to test;
- secondary outcomes: additional planned measures.
This helps prevent researchers from choosing whichever outcome happens to look best after data collection.
Preregistration and Multiple Comparisons
When many tests are performed, false positives become more likely.
Pre-specifying the primary tests helps distinguish:
- planned confirmation;
- exploratory searching.
See Multiple Comparisons.
Preregistration and p-Hacking
p-Hacking can occur when researchers try many analytical choices until a significant result appears.
Preregistration reduces this risk because the original analysis plan is visible.
If researchers later make a different choice, they can explain why.
Transparency matters more than pretending no changes occurred.
Preregistration and HARKing
HARKing means hypothesising after the results are known while presenting the hypothesis as though it came first.
Preregistration makes this harder because the original hypotheses are timestamped.
Exploratory hypotheses should instead be labelled honestly and tested in future studies.
A Worked Example: Learning Study
Researchers compare two study methods.
Before collecting data, they preregister:
- sample size = 200;
- primary outcome = test score after one week;
- exclusion rule = incomplete final test;
- analysis = difference in mean score between groups.
Later, they discover an interesting subgroup pattern.
They can report it as exploratory and recommend replication.
That is good Science.
A Worked Example: Plant Experiment
Students preregister:
- three light conditions;
- height increase after 14 days as primary outcome;
- water and soil as controls;
- mean growth comparison;
- rules for handling dead plants.
This creates a clear record before the results are known.
Registered Reports
A registered report is a publication format where the research question and methods are peer reviewed before data collection.
If the protocol is accepted, the journal commits in principle to publication based on methodological quality rather than whether results are positive.
This reduces publication bias toward exciting findings.
Preregistration Does Not Guarantee Good Science
A preregistered study can still have:
- poor measurements;
- bad sampling;
- confounding;
- underpowered design;
- incorrect analysis.
Preregistration increases transparency. It does not replace scientific quality.
When Plans Need to Change
Unexpected problems happen.
A sensor may fail.
A planned analysis may be inappropriate.
A safety issue may force a design change.
The correct response is:
- make the change;
- document it;
- explain why;
- distinguish planned from unplanned analyses.
Preregistration and Replication
Exploratory findings can become the preregistered hypotheses of the next study.
This creates a healthy cycle:
explore → generate hypothesis → preregister → test → replicate.
See Repeatability and Reproducibility.
Primary Science Foundations
Primary learners can use a simple version:
Before the experiment, write what you plan to change, measure and compare.
Then do not rewrite the plan afterward to make it look as if you predicted the result.
Secondary Science Preregistration
Secondary students should increasingly understand:
- pre-specified hypotheses;
- primary outcomes;
- analysis plans;
- exploratory vs confirmatory research;
- p-hacking;
- registered reports.
How to Practise Preregistration
Before a class experiment, write:
- research question;
- hypothesis;
- primary outcome;
- sample size or repeats;
- analysis plan;
- exclusion rules.
Afterward, compare the actual analysis with the original plan.
Common Preregistration Misconceptions
- preregistration bans exploration;
- preregistered research is automatically correct;
- plans can never change;
- only medical research can preregister;
- exploratory findings are worthless;
- preregistration replaces peer review.
Frequently Asked Questions
What is preregistration?
Preregistration is the timestamped recording of key research questions, hypotheses and analysis decisions before results are known.
Why preregister a study?
To reduce hindsight bias, selective reporting and p-hacking while making confirmatory and exploratory analyses easier to distinguish.
Can researchers change a preregistered plan?
Yes, when justified, but changes should be documented transparently.
What is a registered report?
It is a publication format where the study plan is peer reviewed before data collection and publication is not dependent on obtaining positive results.
Useful eduKateSG Routes
The Core Aim
Preregistration separates prediction from hindsight.
Write the question first. Write the analysis first. Explore freely afterward—but label exploration honestly.
That is the core aim: keep discovery creative while keeping confirmation transparent.
Properly taught kids shine a bright light into the future.