Did you know that children can remember their first prediction differently after they have seen the result? A folded paper bridge holds more coins than a flat one, and suddenly everybody says, “I knew it!” Yet before the test, some learners were convinced the flat sheet would be stronger. An honest experiment needs room for both ideas.
Recording a prediction before collecting results creates a clear record of what the student expected, why they expected it and what evidence could challenge it. The purpose is not to guess the correct answer. It is to practise reasoning before a result is known, then compare a genuine expectation with the observations without rewriting history.
What belongs in a useful prediction?
A prediction names the outcome expected under specific conditions. An explanation gives a reason the student expects that outcome. A hypothesis may describe a broader testable explanation. These can be related, but they are not identical. A pupil can give a worthwhile prediction without pretending to know the mechanism with certainty.
For example: “If two bridges use the same size of paper and cross the same gap, I predict the folded bridge will hold more identical coins before failing.” A possible reason is that folding changes the shape and may make bending harder. The pupil has stated a comparison, an observable result and relevant conditions.
“The folded bridge was stronger because it held nine coins” is not a prediction written after seeing the nine-coin result. It is a description of an observation with an explanation to examine. Labeling the two kinds of sentence accurately matters more than sounding scientific.
Worked classroom example: two paper bridges
Here is a complete fictional activity that families or tutors can use for discussion. A teacher places books the same distance apart, gives pupils equal-sized sheets of the same paper and the same type of coin, and defines failure as the bridge no longer supporting the newest coin for two seconds. One bridge remains flat; one receives lengthwise folds. Coins are added in a consistent position using the same method.
Before the trial, a learner writes: “My prediction: the folded bridge will hold more coins. My reason: its shape may resist bending better. My uncertainty: folds could slip, and one test might not tell us much.” The date and prediction are preserved. Another learner may predict no difference; that is acceptable if their reasoning is clearly recorded.
The observations, recorded separately
- Flat bridge, trial 1: 3 coins before failure.
- Flat bridge, trial 2: 4 coins before failure.
- Flat bridge, trial 3: 4 coins before failure.
- Folded bridge, trial 1: 8 coins before failure.
- Folded bridge, trial 2: 7 coins before failure.
- Folded bridge, trial 3: 9 coins before failure.
The flat bridge’s mean supported count is (3 + 4 + 4) ÷ 3 = 11 ÷ 3 ≈ 3.67 coins. The folded bridge’s mean is (8 + 7 + 9) ÷ 3 = 24 ÷ 3 = 8 coins. The means summarise the three illustrative trials; there is no such thing as supporting a fraction of a coin in one actual test.
These invented numbers are consistent with the learner’s prediction in this specific setup. They do not prove that every folded paper bridge is stronger than every flat bridge. Fold design, paper weight, support distance, coin placement and how failure is defined could change the comparison.
The four-column record that prevents hindsight
- Before: prediction with date, relevant conditions and a reason, including what result would surprise the student.
- During: unchanged raw readings for each trial, plus any procedure changes or interruptions.
- After: a precise comparison between the prediction and the collected observations; use quantities and units where relevant.
- Next: a better question, repeatable test or alternative explanation suggested by the evidence.
For the bridges, the final note could say: “My prediction was supported in the three trials using our materials. I did not expect the folded bridge to vary from seven to nine coins. I would like to repeat with another sheet and the same loading position.” That paragraph shows intellectual honesty and curiosity, not just a correct guess.
Diagnose the skill before running another experiment
Ask a child to distinguish three statements: “I think the folded bridge will hold more coins”; “It held eight coins in trial one”; and “Folding makes the paper resist bending.” The first is a prediction, the second an observation, and the third a possible explanation that needs appropriate evidence. If the child mixes them up, work on this classification before increasing the complexity of the experiment.
Next give a new comparison: “Will a ball roll farther down a smooth or a rough track under otherwise similar conditions?” Ask for a prediction before the test, a short reason and one condition to keep comparable. Do not reveal the answer. After trying it safely, the student should read the original prediction verbatim and discuss the observation without changing the prediction sentence.
Why a wrong prediction can be educationally valuable
A prediction can be incorrect while the reasoning process is thoughtful. Science uses unexpected results to improve explanations, check procedures and pose new questions. If a teacher only praises pupils whose guesses came true, learners may learn to hide uncertainty or change their records after the fact. A more useful response is “What made you expect that? What did we observe? What should we check next?”
The prediction should be specific enough to test, but it should not be framed as a promise. “I predict A will support more coins than B” is clearer than “A is the best design in the world.” A classroom investigation usually has a narrow method and a small number of trials; conclusions need the same modest scope.
What students, parents and tutors can do tomorrow
- Student: write one prediction before touching the equipment. Take a photo of the notebook page if allowed, or simply keep its dated entry.
- Parent: ask “What do you expect, and what could change your mind?” Avoid asking the child to guess what the adult wants to hear.
- Tutor: show a worked example of prediction versus observation, then let the learner classify new statements independently.
- Teacher: collect predictions before demonstrations begin and evaluate the reasoning separately from whether a guess was borne out.
- Everyone: record a changed procedure honestly rather than editing the original prediction to fit the final outcome.
Common mistakes and sensible limits
- Writing the prediction only after the results have become known.
- Calling the explanation and the measurement the same thing.
- Changing two important features at once, then claiming that one alone caused the difference.
- Reporting only the trial that matched the prediction and concealing the others.
- Treating an unexpected result as a personal failure instead of an invitation to inspect the method.
The paper-bridge activity is not a formal strength test for building materials. Different paper types, folds, coin placement and classroom handling can alter its results. Some demonstrations cannot be safely repeated at home; follow teacher guidance, use age-appropriate materials and never improvise with hazardous substances or electrical equipment.
Frequently asked questions
Must a child predict the exact number of coins?
No. A directional prediction such as “more than the flat bridge” can be testable. Older learners may add a number or range if they have a reason for doing so.
What if the prediction was partly right?
State exactly which part agreed with the observations, which part did not and whether the test was precise enough to decide. Do not force a yes-or-no verdict when the evidence is mixed.
Should the student change the prediction after a failed trial?
Keep the original as the record of the first test. Write a new dated prediction for the next version, explaining what evidence prompted the change.
The lesson is better thinking, not better guessing
One dated prediction, a fair comparison, raw observations and an honest reflection form a small but powerful learning cycle. A child who can explain why a prediction changed is making progress even when the first guess was wrong. Build the cycle with the Core Aim guide to pilot-testing a science investigation, the guide to honest scientific limitations and the wider eduKate Education Hub.
For a procedural reference, Science Buddies’ conducting-an-experiment guide recommends careful raw records and noting changes to procedures. The diagnosis–explanation–practice–independent-check progression also informs eduKate’s Secondary 1 Mathematics learning benchmark.