Asking a research question
Science starts with a question that we can test by measuring something. A good question names two things and how they may be linked.
- Weak: 'Why is sugar nice?' (cannot be measured)
- Better: 'Does the water temperature change how fast sugar dissolves?'
Then write a hypothesis: an idea that can be proved wrong. 'The warmer the water, the less time it takes.' Say why you think so (hot particles move faster). After the test, the result may support the hypothesis or not. Both are useful.
The experimental method is the loop: question → hypothesis → plan → test → results → conclusion → new question.
Planning an experiment: the three kinds of variable
A variable is anything that can change in the test.
- Independent variable: the one thing you change on purpose (water temperature).
- Dependent variable: the one thing you measure (time to dissolve).
- Controlled variables: everything else, kept the same (amount of sugar, volume of water, stirring, type of glass).
This is a fair test. If two things change at once, you cannot tell which one caused the result.
A plan lists: the question, the variables, the apparatus, the steps in order, the range of values (for example 20, 40, 60 °C), how many repeats, and the safety rules (see the lab-safety lesson). Always plan how you will record before you start.
Measurement and uncertainty
Choose the right instrument and use it well: a ruler for length, a measuring cylinder for volume, a balance for mass, a thermometer for temperature, a stopwatch for time. Read the scale at eye level, start from zero, and write the unit.
No measurement is perfect. Uncertainty tells how far the true value might be from your reading.
- Instrument limit: a ruler with millimetre marks can be read to about ± 0.5 mm.
- Repeat readings: take at least 3 to 5. Mean = sum ÷ number of readings.
- Spread (simple uncertainty): ± half the range = (largest − smallest) ÷ 2.
- Percentage uncertainty = (uncertainty ÷ mean) × 100%.
Accurate means close to the true value; precise means the repeated readings are close to each other. A random error makes readings scatter (reaction time on a stopwatch); a systematic error pushes all readings the same way (a balance that does not start at zero). Repeating reduces random errors, but not systematic ones, so check your instrument first.
Lab notebook and reporting
A lab notebook is your honest record, written while you work, in pen, with the date.
- Title, date and aim.
- Apparatus and a short method (a labelled sketch helps).
- A results table with headings and units: Temperature (°C) | Time 1 (s) | Time 2 (s) | Time 3 (s) | Mean (s).
- Observations: what you saw, smelt (never taste) or heard, including surprises and mistakes. Cross out errors with one line; never erase.
The report turns this into a clear story for others:
- Aim and hypothesis
- Method (variables, apparatus, steps, safety)
- Results (table and a graph with labelled axes and units)
- Conclusion that answers the question using your data
- Evaluation: errors, what you would improve, what to test next
Lab projects
A lab project is a longer investigation that you plan yourself, often in a team.
- Pick a real question and check it is safe and possible with the time and tools you have.
- Read a little first so you do not repeat what is already known.
- Write a plan and get it approved by the teacher.
- Share jobs (measurer, recorder, timekeeper) and rotate them.
- Collect data, check it as you go, and repeat anything odd.
- Present with a short talk or poster: question, method, graph, conclusion, what you learned.
Keep a project diary and give credit to anyone whose help or data you used.
Science in everyday life
The same skills help you decide in daily life.
- Which detergent removes a stain best? Fair test with the same cloth, stain, water and time.
- Does the cooling of a pot depend on its lid? Measure temperature every minute.
- Which phone battery setting lasts longer? Repeat, compare means.
- Is a claim on an advert true? Ask: what was compared, how many times, and was it fair?
Thinking like this protects you from weak claims and helps you check news and health advice with care.
Key formulas and definitions
- Mean = (sum of readings) ÷ (number of readings)
- Uncertainty ≈ ± (largest − smallest) ÷ 2
- Percentage uncertainty = (uncertainty ÷ mean) × 100%
- Fair test: change 1 variable, measure 1 variable, control the rest
- Rate = change ÷ time (for example mass dissolved per second)
Worked examples
1. Five dissolving times at 40 °C: 31, 29, 30, 32, 28 s. Find the mean.
Sum = 31 + 29 + 30 + 32 + 28 = 150. Mean = 150 ÷ 5 = 30 s.
2. For the same readings, find the uncertainty (half the range).
Range = 32 − 28 = 4 s. Uncertainty = 4 ÷ 2 = ± 2 s. We write the result as 30 ± 2 s.
3. A plant-growth test: pot A gets sunlight, pot B is kept in a cupboard. Both have the same soil, water and seeds. Name the variables.
Independent: light. Dependent: height of the plant (or growth). Controlled: soil, water amount, seed type, pot size, temperature.
4. A length is 12.0 cm with an uncertainty of ± 0.1 cm. What is the percentage uncertainty?
(0.1 ÷ 12.0) × 100% = 0.83%, about 0.8%.
5. A balance reads 0.4 g when nothing is on it. You measure 25.0 g of salt. What is the true mass, and what kind of error was this?
True mass = 25.0 − 0.4 = 24.6 g. This is a systematic error (zero error), which every reading shares. Repeating does not remove it; you must correct it or reset the zero.
Common mistakes
- Changing more than one thing at a time, so the test is not fair.
- Taking one reading and trusting it. Repeat at least three times.
- Writing the conclusion before looking at the data, or ignoring results that do not fit.
- Leaving out units, headings or axis labels in tables and graphs.