KEEP IT FAIR
Change one thing
Decide what you are testing and keep everything else the same: amounts, sizes, timing, place.
Try the fair test checkKITCHEN-TABLE SCIENCE
Change one thing.
Keep the rest.
A fair test is the simplest habit real labs share, and you can run one at home.
Which kitchen paper soaks up the most? Do beans sprout faster in the light? Results only mean something if the test is fair, repeated and written down. Here is how scientists make sure, in plain English.
Check your test Read the short answerPlain English, with the sources linked.
01 Change one thing02 Repeat it03 Judge it blindWrite it all down
THE SHORT ANSWER
Fair, repeated, recorded.
Three habits separate a real result
from a lucky one.
MAKE IT HOLD
Repeat it
One try can be a fluke. Repeating a test shows whether the result is real and how much it varies.
See what repeats revealed
KEEP A RECORD
Write it down
Date it, note what, why and how, and keep the results you did not expect as well as the ones you did.
How labs keep notesA QUICK FAIR TEST CHECK
Which kitchen paper
soaks up the most?
You want to compare three brands of kitchen paper. Pick how you would run the test and see how it rates.
The ratings follow the principles in England’s primary science curriculum and the research linked below. They are a guide, not a marking scheme.
TWO STUDIES WORTH KNOWING
Why labs repeat
and judge blind.
Even professional results can shrink
when the test is run again, or judged blind.
REPEAT IT · 2015
Significant results, original vs repeat
97% vs 36%
A team repeated 100 published psychology studies. Almost all the originals had reported a significant result; only about a third of the repeats did, and the effects came out about half as large.
JUDGE IT BLIND · 2012
Average exaggeration by judges who knew
36%
Across 21 medical trials, assessors who knew which treatment each patient had received reported effects 36% larger on average than assessors who did not know.
Neither study means science is broken. They show why scientists build repeats and blinding into their tests, and why a home test is more convincing when it does the same.
LAB FACTS
Five things labs
know by heart.
01
Fair tests start at primary school
England’s national curriculum has pupils setting up comparative and fair tests in years 3 and 4, and recognising and controlling variables by years 5 and 6.
02
Effects often shrink on a second try
In the 2015 project that repeated 100 psychology studies, the repeated effects were on average about half the size of the originals.
03
Ink, never whiteout
NIH record-keeping guidance asks for entries in ink, signed and dated, with mistakes lined out and dated rather than erased or covered with whiteout.
04
Write for a stranger, years later
One University of Pennsylvania lab tells its members to record enough for others to reproduce the results, as the notes may be needed 7 to 10 years after they leave.
05
Scientists worry about it too
In a 2017 survey of Nature journal authors, 36% of those who answered said there was a significant crisis of reproducibility, rising to 41% in the biomedical and life sciences.
TRY IT AT HOME
Four lab habits
for the kitchen table.
- 01
Write the plan first
Before you start, note the question, the one thing you will change and what you will measure. Labs plan before they begin, too.
- 02
List what stays the same
Same amount of water, same size of sheet, same spot on the windowsill, same timing. Each one you hold steady makes the test fairer.
- 03
Run it more than once
Repeat each version and compare the results side by side. If they jump about, the difference you saw may be luck.
- 04
Let someone else judge
Hide which is which and ask a helper to measure or pick the winner. Then write everything down, surprises included.
QUICK QUESTIONS
Fair tests, answered.
Short answers to the questions people ask most.
What is a fair test?
A fair test changes only the one thing you want to find out about and keeps everything else the same, so that any difference you see can be put down to that one change. In England, pupils start setting up fair tests in years 3 and 4 of primary school.
What is the difference between a comparative test and a fair test?
A comparative test simply compares, such as a plant kept in light against one kept in the dark. A fair test goes further by recognising and controlling the other variables. England’s national curriculum expects this by years 5 and 6.
Why repeat an experiment if it already worked?
Because one result can be luck. When researchers repeated 100 published psychology studies in 2015, 97% of the originals had reported a significant result but only 36% of the repeats did, and the effects were on average about half as large.
Why should the person measuring not know which is which?
Expectations can nudge judgement without anyone meaning to. A 2012 analysis of 21 medical trials found that assessors who knew which treatment patients had received exaggerated the effect by 36% on average compared with assessors who did not know.
How should I write up a home experiment?
Date each entry and note what you did, why, how and what happened, including results you did not expect. Professional guidance asks for ink, a line through mistakes rather than erasing them, and enough detail for someone else to repeat the work.
SOURCES
Where the facts come from.
A NAME WITH ROOM TO EXPERIMENT
doodlelab.com could be yours.
doodlelab.com is part of the DoodleSale portfolio. Check the current listing and available purchase or lease options.