The core idea
Science connects explanations with observations that other people can check. A useful investigation states a question, makes a fair comparison, records what happened and explains both what the evidence supports and what remains unanswered.
1. Observation, hypothesis and prediction
An observation is something you notice or measure: a wet cloth in sunlight became dry before a similar cloth in shade. A hypothesis proposes an explanation: the warmer conditions increased evaporation. A prediction states what you expect if that explanation is useful: under otherwise similar conditions, the warmer cloth should lose water faster. These are three different jobs. The observation describes an event; the hypothesis tries to explain it; the prediction gives you something to check.
One comparison does not settle the cause. The sunny position may also have been windier. That alternative matters because moving air can affect drying too. Scientific reasoning improves an explanation by asking which observations could distinguish it from another plausible explanation. A confident sentence without such a check is still a claim.
Sources: NCERT Curiosity: The Wonderful World of Science ↗ · NCERT: Nature of Science, Grade 9 syllabus ↗
2. What is a variable, and why control it?
A variable is a condition or quantity that can take different values. In a study of stirring and dissolving sugar, the independent variable is the condition deliberately changed: stirred or not stirred. The dependent variable is the measured result: the time until no crystals are visible. Controlled variables are conditions kept alike, such as the starting water temperature, water volume, sugar mass and cup shape. Controlling them makes a comparison easier to interpret.
Suppose cup A has warm water and stirring, while cup B has cool water and no stirring. If A finishes first, the design cannot separate temperature from stirring. This is a confounded comparison: two possible causes changed together. Improve it by using equal starting temperatures and changing only stirring. A control is a comparison condition, not a magical guarantee that every hidden influence has disappeared.
3. A measurement needs a rule and a unit
Before starting, decide what counts as the endpoint. “Looks dissolved” is ambiguous if one person stops at the last large crystal and another waits for complete clarity. Use the same stated rule for every trial. Write seconds rather than just “40”, and record how much water and sugar you used. Someone else needs these details to repeat the comparison rather than merely imitate its appearance.
Precision concerns how closely repeated measurements agree under the same conditions. Resolution is the smallest change an instrument can distinguish; accuracy concerns closeness to the appropriate reference value. A clock that is always five minutes fast can give very consistent readings while being inaccurate. Repeating a measurement may reveal random variation, but does not automatically fix a consistently wrong instrument. Recording many decimal places also cannot create accuracy the instrument lacks.
Sources: NCERT: Nature of Science, Grade 9 syllabus ↗ · NIST: measurement precision and accuracy ↗
4. Worked example: three trials tell a fuller story
Consider these invented practice results, not an actual experiment: stirred water takes 38, 42 and 40 seconds; unstirred water takes 96, 104 and 100 seconds. The mean stirred time is (38 + 42 + 40) ÷ 3 = 40 seconds. The unstirred mean is 100 seconds. In this example, stirring shortened the average time by 60 seconds. Expressed relative to the unstirred mean, that is 60 ÷ 100 × 100 = 60% less time.
Keep the individual results as well as the mean. A value of 400 seconds in a fourth trial deserves investigation: perhaps the timer was left running, or a different sugar amount was used. Do not remove it merely because it spoils the pattern. Record a known mistake and explain how you handled it. If the reason is unknown, report the unusual value and repeat the test carefully.
An evidence-building cycle
- ObserveDescribe what you notice.
- QuestionAsk something you can investigate.
- TestCompare, measure and record.
- ReviewCheck whether the evidence fits.
5. Association does not, by itself, establish a cause
Imagine that ice-cream sales and electricity use both rise during hotter weeks. Buying ice cream does not necessarily cause the electricity use. Temperature can influence both. This third factor is a possible confounder. A pattern can suggest a question worth investigating, but the causal explanation needs additional evidence and a plausible mechanism. The same caution applies when a student studies more and receives better marks: sleep, prior knowledge and question difficulty may also matter.
Not every science question permits a controlled experiment. Astronomers cannot rearrange stars to test a theory; they compare predictions with observations. Ecologists may combine field measurements, comparisons and models. What matters is whether the method can answer the particular question, how alternative explanations were considered and whether independent evidence points in the same direction.
Sources: NCERT Curiosity: The Wonderful World of Science ↗ · NCERT: Nature of Science, Grade 9 syllabus ↗
6. Models explain selected features of the world
A model is a simplified representation used to explain or predict. A map leaves out individual stones yet helps you plan a route. Scientific models likewise make some features central and leave others aside. An atomic model is useful for particular questions; it is not a tiny photograph of an atom. When new evidence exposes a limitation, a model can be refined or replaced without implying that every earlier calculation was useless.
A good conclusion is proportional to the evidence. For the sugar example, say that stirring reduced dissolving time under the tested conditions. Do not claim that every substance behaves identically or that the test proves a health benefit. The complete argument has three parts: the claim, the measurements supporting it and the reasoning connecting those measurements to the claim.
PUT IT INTO PRACTICE
Analyse a fair-test plan
- A learner compares 100 ml warm water with stirring against 200 ml cool water without stirring. List the variables that changed, then design a fairer comparison on paper.
- For practice results 30, 36 and 33 seconds, calculate the mean. Write the endpoint rule and one possible source of measurement error.
- Explain your answer: the mean is 33 seconds; water volume and temperature should be matched when testing stirring. Compare this explanation with yours and revise any unsupported conclusion.
Check your understanding
What makes the warm-water/stirring comparison unfair?
Temperature and stirring change together, so either could explain the result. Match temperature and other relevant conditions, then compare stirring conditions.
Does a mean replace the original measurements?
No. The mean summarises the values, while individual readings show variation and unusual results. Both are needed to understand the investigation.
If two things rise together, has one caused the other?
Not necessarily. A third factor can affect both. The hotter-week example links temperature to ice-cream sales and electricity use without proving that sales cause electricity demand.
Why can a changing scientific explanation be a strength?
A revision can bring the explanation into better agreement with reliable evidence. The change should be justified by observations and reasoning, not made simply to protect a favourite idea.