5.2 Controls, Repetition & Reliability

What Is a Control?

A control (or control group, or control setup) is the version of an experiment that is left in normal, untreated conditions, so you have a baseline to compare against. If you are testing whether fertiliser helps plants grow, the control is the plant given no fertiliser at all. If you are testing whether moisture causes mould, the control is the slice of bread with no added water.

Why bother? Because things change on their own. Plants grow a little even without fertiliser, and bread sometimes grows mould anyway. Without a control, you cannot tell whether your treatment caused the change or whether it would have happened regardless. The function of a control is to strip away that doubt: any extra change in the test setups, beyond what happened in the control, can be credited to the independent variable. ICAS Paper C asks directly about the function of controls in biological experiments, so be ready to say it in your own words: the control shows what happens without the thing being tested, so the results can be compared.

Control vs Controlled Variables: The Trap ICAS Sets

These two phrases sound almost identical, and ICAS knows it. Keep them straight:

A memory hook: the control is a thing (a setup or a group), while a controlled variable is a rule (something you hold constant everywhere). A single experiment has one control setup but many controlled variables. When a question asks "what is the control in this experiment?", look for the untreated setup - not for a list of conditions.

Repetition, Replication and Sample Size

Even a fair test can produce a fluke. Three ideas protect against that:

• Repetition means repeating your own trials several times and comparing (or averaging) the results. If three trials agree and one is wildly different, the odd one was probably an error, not a discovery.
• Replication means other people repeat the whole experiment independently. Scientists only trust surprising results once they have been replicated.
• Sample size is how many individuals or items you test. One fertilised plant might be naturally taller than one unfertilised plant, fertiliser or not - living things vary. Testing ten plants in each group smooths out that natural variation and makes a real difference visible.

Here is repetition in action. Jayden drops a ball three times and measures the bounce height: 58 cm, 61 cm and 59 cm. Those results cluster together, so he can trust the bounce is about 59 cm. His friend drops the same ball once from a ladder and measures 84 cm - one trial, no comparison, no way to know whether the drop was clumsy or the ball truly bounces higher from greater heights. More trials turn a guess into evidence, and taking an average (adding the results and dividing by the number of trials) is the usual way to summarise repeated measurements.

Reliable vs Valid

Two words ICAS uses at the upper year levels, in plain language:

• A result is reliable when repeating the experiment gives roughly the same answer each time. Consistency is the key word.
• An experiment is valid when it actually tests what it claims to test - a fair test measuring the right thing.

Think of a bathroom scale that reads 2 kg too heavy every single time. It is reliable (same reading every time) but not valid (the reading is wrong). An experiment can also be repeatable yet invalid if an uncontrolled variable is really causing the results.

Random Sampling, Briefly

Sometimes you cannot test every member of a group - every daisy in a field, every ant on the school oval. Instead you take a sample (a smaller group chosen to represent the whole), and to keep it unbiased you choose at random, for example by tossing a square quadrat without aiming. Choosing randomly stops you from accidentally picking only the shady spots or only the tall plants. Sampling methods are covered fully in the ecology chapter; here, just know that how you choose your sample is a design decision ICAS can ask about.

Ethics and Safety in Investigations

Real investigations involve real materials and sometimes living creatures, and scientists have responsibilities:

• Safety first: wear safety glasses when heating or mixing, never taste anything in an experiment, wash hands afterwards, and follow teacher instructions - and say so when describing a method.
• Treat living subjects fairly: cause as little harm or stress as possible. Return insects and small animals to where they were found, keep them only briefly, provide food and water while they are observed, and use plants instead of animals when a plant can answer the question. Paper J includes questions on the ethics of the use of living subjects in experiments, and the expected answer is always the humane, minimal-harm choice.

A well-designed investigation respects its equipment, its data and its living subjects - and ICAS rewards students who show they understand all three.