Investigating
What you must understand
An investigation tests an idea with a fair test. Identify the aim, the independent variable you deliberately change, the dependent variable you measure, and the controlled variables kept constant. A control group receives no treatment or a standard treatment, so the effect of the change can be judged against a baseline. Plan how each variable will actually be measured (the operational definition), choose appropriate instruments, decide on sample size and repetition, and record results in a clearly headed table with units. Consider reliability (consistency of repeats), validity (whether the test really measures what it claims), safety and ethics. Procedures must be written so another person could repeat them exactly.
Traps to watch
- Changing more than one variable, then attributing the result to one of them.
- Confusing "the control group" with "the controlled variables".
- Testing one plant per treatment and treating the result as general.
- Measuring a variable in a way that does not match the aim.
- Forgetting to keep conditions such as temperature, light or time identical.
Worked example
To test whether fertiliser affects plant growth: independent variable is the mass of fertiliser added; dependent variable is plant height measured with a ruler to the nearest millimetre at the same time each week; controlled variables include plant species, soil type, pot size, water volume, light and temperature; a zero-fertiliser group is the control; use ten plants per treatment and average the results. Rule of thumb: change one thing, measure one thing, keep everything else the same.