Good science begins with good observation. Observing means using your senses carefully and honestly: looking, listening, smelling, touching and sometimes tasting when it is safe. A casual look at a leaf tells you it is green. A careful observation tells you it is dark green on top, paler underneath, has a wavy edge, a thick central vein, and a length of 12 centimetres. Scientists write these observations down at once, because memory changes and fades. A science notebook with the date, a title, a labelled drawing, measurements and a short conclusion is the most important tool you will own this year. Drawings do not need to be beautiful; they need to be accurate and labelled.
When we test an idea, the test must be fair. Suppose you want to know whether plants grow faster with more water. If you give one plant more water but also more sunlight and better soil, you cannot tell which thing made the difference. A fair test changes only one variable at a time and keeps everything else the same. The thing you change is the independent variable, the thing you measure is the dependent variable, and the things you keep the same are the controlled variables. Scientists also repeat their tests, because one result could be an accident. They use tables to record numbers and graphs to see patterns. A good conclusion says what the evidence shows and admits what is still unknown. This is the way the NCERT Curiosity book asks you to work all year: observe, question, test fairly, record, and share.
Example: Test whether salt makes water evaporate more slowly. Take two identical small dishes and pour exactly 50 millilitres of water into each. Add two spoons of salt to one dish only. Place both dishes side by side in the same sunny spot. Every day at the same time, mark the water level or measure the depth with a ruler and record it in a table. Expected result: the plain water disappears first, while the salty dish keeps water longer and finally leaves white salt crystals. Write down your independent variable, dependent variable and at least three controlled variables.