Ready to Test Your Ideas Like a Scientist?
Why does sugar disappear in seconds in hot milk tea, but sit at the bottom of a glass of iced lemon tea? Scientists ask questions like this, and then they test their ideas carefully.
In this chapter you will learn how to write a hypothesis, identify variables, set up a controlled experiment, decide whether results support or refute a hypothesis, and check how reliable a conclusion is. These skills help in every science topic!
1. What Is a Hypothesis?
A hypothesis is a statement that describes, predicts or explains something we observe in nature. Most importantly, it can be tested by a scientific investigation.
It is not a wild guess, but an educated idea based on what you have observed or already know.
Three jobs a hypothesis can do
Describe – "Sugar dissolves faster in hot water than in cold water."
Predict – "If I use water at 80 °C instead of 20 °C, the sugar will dissolve in a shorter time."
Explain – "Sugar dissolves faster in hot water because the water particles move faster and separate the sugar particles from the crystals more quickly."
What makes a good hypothesis?
1. It is testable – you can do an experiment or collect observations to check it. "Blue is the best colour" is an opinion, so it cannot be tested.
2. It is clear and specific – it says what will be changed and what will be measured.
3. Results could show it is wrong – if no result could ever disagree with it, it is not useful.
Memory aid: Use the pattern "If … then … because …".
From observation to hypothesis
Observation: In a cha chaan teng, sugar stirred into hot milk tea disappears quickly, but sugar in iced lemon tea often stays at the bottom. That is one reason iced drinks often come with sugar syrup.
Question: Does the temperature of water affect how fast sugar dissolves?
Hypothesis: If the temperature of the water is higher, then sugar will dissolve in a shorter time.
Key takeaway: A hypothesis is a testable statement that describes, predicts or explains. Start from an observation, ask a question, then write your hypothesis.
2. Quick Recap: Three Types of Investigation
You met these in the chapter Practice of science:
Fair testing – change one factor, keep the others the same and see what happens, e.g. does water temperature affect how fast sugar dissolves?
Classifying – sort things into groups using features you can observe or test, e.g. sorting materials into electrical conductors and insulators.
Pattern seeking – look for patterns or links in data when you cannot control every factor, e.g. do taller students have larger hand spans?
Key takeaway: This chapter focuses mainly on fair tests.
3. Variables: Change, Measure, Keep the Same
A variable is any factor that can change or be changed in an investigation. There are three types.
Independent variable – the one you CHANGE
The independent variable is the factor you deliberately change to see its effect. In a fair test, you change only one independent variable.
Dependent variable – the one you MEASURE
The dependent variable is the factor you measure or observe. It is called "dependent" because it depends on the independent variable.
Control variables – the ones you KEEP THE SAME
Control variables are all the other factors that could affect the result. You keep them the same so that the test is fair.
Memory aid: "I change the Independent; the Dependent Depends on it; the Controls stay Constant."
Example: growing bean seedlings
Question: Does the amount of light affect how tall bean seedlings grow?
Independent variable: amount of light (e.g. hours of light each day)
Dependent variable: height of the seedlings after 7 days
Control variables: type of bean, volume of water, soil, temperature, pot size
Try it yourself
Does the drop height of a ball affect how high it bounces?
Answer: independent = drop height; dependent = bounce height; control = same ball, same floor surface.
Key takeaway: Independent = what you change. Dependent = what you measure. Control = what you keep the same.
4. Controlled Experiments: Finding Cause and Effect
What is a controlled experiment?
A controlled experiment changes only the independent variable and keeps all control variables the same. Many have two parts:
Experimental set-up – has the factor being tested.
Control set-up – exactly the same, except that the factor being tested is missing (or kept at its normal level). It gives us something to compare with.
Example: does fertiliser help plants grow?
Use two groups of the same kind of bean seedlings, with the same soil, light and temperature. Give the experimental group fertiliser solution and the control group the same volume of plain water. If the fertiliser group grows taller, the fertiliser caused it, because it was the only difference.
Confounding factors
A confounding factor is a factor other than the independent variable that also changes and affects the result. We can then no longer tell what caused the change.
Example: Ming dissolves sugar in hot water and stirs it, but leaves the sugar in cold water unstirred. The sugar in hot water dissolves faster, but was it the temperature or the stirring? He cannot tell! Stirring is a confounding factor.
Why controlled experiments show cause and effect
When the independent variable is the only difference between the set-ups, any change in the dependent variable must be caused by it. This is how a controlled experiment helps us establish causality (cause and effect).
Without control, two things can happen together even if one does not cause the other. On hot days, more ice cream is sold and more people swim at Hong Kong's beaches, but ice cream does not make people swim. Hot weather affects both.
Key takeaway: Change only one variable, keep the rest the same and use a control set-up. This removes confounding factors so you can claim cause and effect.
5. Conclusions: Support or Refute?
After collecting results, you compare them with your hypothesis.
If the results agree with the hypothesis, the conclusion supports the hypothesis.
If the results disagree, the conclusion refutes (goes against) the hypothesis. You can then revise it and test again.
Why we never say "proved"
One experiment can never prove that a hypothesis is true forever. You only tested certain conditions, and new evidence might show something different. So we say the evidence supports a hypothesis.
A refuted hypothesis is not a failure. It still teaches us something.
Did you know?
About 400 years ago, Jan Baptista van Helmont grew a small willow tree in a pot of weighed soil for five years, giving it only water. The tree gained a lot of mass, but the soil lost only a tiny amount. This went against the old idea that plants are made mainly from soil. He concluded the mass came from water, but later scientists found that much of it comes from carbon dioxide in the air!
Key takeaway: Results support or refute a hypothesis. They never "prove" it.
6. Is the Conclusion Reliable?
A reliable conclusion is one we can trust, because we would get similar results if the investigation were done again. There are two ways to check.
Repeat the experiment
Repeating means doing the same experiment again with the same set of apparatus, usually three times. Results that are close together are consistent. Repeating also helps you spot an unusual result (an anomaly) and calculate an average:
\(\text{average} = \dfrac{\text{sum of results}}{\text{number of trials}}\)
Reproduce the experiment
Reproducing means doing the same investigation in a different environment or with a different set of apparatus, often by another group, e.g. another class in another laboratory using a digital thermometer. If they get similar results, we can be much more confident in the conclusion.
Key takeaway: Repeat with the same apparatus to check consistency. Reproduce in a different setting or with different apparatus to check the conclusion still holds.
7. Worked Example: Temperature and Dissolving Sugar
Step 1 – Hypothesis: If the water temperature is higher, then sugar will dissolve in a shorter time, because particles in hotter water move faster.
Step 2 – Variables:
Independent: water temperature (20 °C, 40 °C, 60 °C, 80 °C)
Dependent: time for the sugar to dissolve completely (in seconds)
Control: type and mass of sugar (5 g), volume of water (100 mL), beaker size, stirring speed
Step 3 – Method: Warm 100 mL of water in a beaker to the chosen temperature (check with a thermometer). Add 5 g of sugar, start the stopwatch and stir steadily. Stop when no crystals can be seen. Do each temperature three times. Take care with hot water.
Step 4 – Example results (time in seconds):
20 °C: 92, 88, 90 – average 90
40 °C: 61, 58, 61 – average 60
60 °C: 41, 39, 40 – average 40
80 °C: 26, 24, 25 – average 25
For example, at 20 °C: \(\text{average} = \dfrac{92 + 88 + 90}{3} = 90\ \text{s}\)
Step 5 – Conclusion: As the water temperature increased, the time for the sugar to dissolve decreased. The results support the hypothesis.
Step 6 – Reliability: The three trials at each temperature are close together, so the results are consistent. Another group could reproduce the experiment with a different thermometer and stopwatch.
Key takeaway: A good investigation goes: observation → hypothesis → variables → controlled method → results → conclusion → reliability check.
Common Mistakes to Avoid
Common mistake 1: Writing "My experiment proved my hypothesis." Instead, write "The results support my hypothesis."
Common mistake 2: Mixing up control variables and the control set-up. Control variables are the factors you keep the same. The control set-up is a separate set-up used for comparison.
Common mistake 3: Changing two things at once (e.g. temperature and stirring). You cannot tell which one caused the result.
Common mistake 4: Mixing up independent and dependent variables. Ask "Which one do I change?" (independent) and "Which one do I measure?" (dependent).
Quick Review
1. A hypothesis describes, predicts or explains a phenomenon, and it must be testable.
2. The three types of investigation are fair testing, classifying and pattern seeking.
3. Independent variable = changed; dependent variable = measured; control variables = kept the same.
4. A controlled experiment with a control set-up removes confounding factors, so we can identify cause and effect.
5. A conclusion supports or refutes a hypothesis. It never proves it.
6. Check reliability by repeating (same apparatus) and reproducing (different environment or apparatus).
Great work! Next time you see something puzzling, ask a question, write a hypothesis and test it!