Listing controlled variables in your plan is easy. Actually holding them steady while you work is the hard part — and it is the difference between data that answers your question and data that answers a question you never asked.
📚 What you need to know
Controlling variables makes your results valid: it means any change in the DV really was caused by the IV.
It mainly targets systematic error — error that pushes every reading in the same direction.
Calibration means checking an instrument against a known standard and adjusting it: pH meters against buffers, thermometers against ice and boiling water, balances zeroed before use.
An eyepiece graticule must be calibrated against a stage micrometer at each magnification.
Keep conditions constant with a water bath or incubator, and let samples acclimatise before starting.
If you cannot control something, monitor and record it, then discuss it in your evaluation.
Living material varies, so use similar samples, replicates and, in fieldwork, random sampling.
A control run has everything except the independent variable, and proves the IV caused the effect.
Random error and systematic error
Two different problems, fixed in two different ways.
Random error scatters readings either side of the true value — slight differences in timing, in judging a colour, in how much water clings to a potato cylinder. Repeats and means reduce it.
Systematic error shifts every reading the same way. A thermometer that reads 2 °C high, a balance that was never zeroed, surface water left on tissue before weighing. Repeating the experiment does not help at all — you just get the same wrong answer more precisely. Calibration and careful technique are what fix it.
Why this distinction matters. If your evaluation says “the results could be improved by doing more repeats”, you have only addressed random error. Examiners want to see that you can spot a systematic problem too — and those are usually the ones that made your numbers wrong rather than just messy.
Calibration
Calibration is checking an instrument’s readings against something you know is right, and adjusting it if needed. It takes a few minutes and protects every number you collect afterwards.
The same logic applies to a pH meter: check it at two known buffers, and readings between them can be trusted.
Instrument
How to calibrate or check it
What goes wrong if you skip it
pH meter
Calibrate against at least two standard buffers, e.g. pH 4.00 and pH 7.00, before taking readings
Every pH value is shifted, so your buffers are not the pH you think they are
Digital thermometer or temperature probe
Check in crushed melting ice (0.0 °C) and boiling distilled water (100.0 °C at standard pressure)
Water bath temperatures are all wrong by the same amount
Oxygen or carbon dioxide sensor
Calibrate against the known concentration of that gas in air
Rates of photosynthesis or respiration are consistently over- or under-read
Eyepiece graticule
Calibrate against a stage micrometer, separately at each magnification
Every cell measurement is wrong, and changing magnification changes the error
Digital balance
Zero (tare) it before every reading, with the container in place
You weigh the container as well as the sample, every time
The graticule one catches people out. The graticule divisions do not change, but what each division represents changes completely when you switch objective lens. Recalibrate at every magnification you use, and say so in your method.
Keeping conditions constant
Living things respond to their surroundings, so the surroundings have to be pinned down.
Temperature is usually the most important. A thermostatically controlled water bath or an incubator holds it steady far better than a room does.
Acclimatise your samples. Put everything into the controlled conditions for a set time — ten minutes is typical — before you mix or start timing. Otherwise the reaction starts while the tube is still warming up.
Light is controlled by fixing the distance from a lamp and keeping it constant, or by working away from windows.
Humidity, pH and volume matter in specific investigations; decide which ones plausibly affect your DV.
Some things genuinely cannot be held constant. Room temperature drifts through a lesson. In that case, do the next best thing: measure and record it at intervals, then discuss the possible effect in your evaluation. A recorded, acknowledged variable is far better than an ignored one.
Samples and biological variation
No two potatoes are the same, and no two leaves from the same plant are identical. That variation is real biology, not sloppiness — but it can drown your trend if you let it.
Make samples as alike as possible. Same variety, same plant, same age, same size. Where sex or age could matter in an animal study, match those too.
Use replicates. Five cylinders at each concentration rather than one means a single odd sample cannot swing the mean.
Sample randomly in the field. In ecological work, use random coordinates to place quadrats rather than choosing where to put them. Choosing “a good spot” is bias, even when you do not mean it to be.
Representative means the sample behaves like the whole. The bigger and more random the sample, the more likely that is. A single quadrat in an interesting-looking patch tells you about that patch and nothing else.
The control run
A control is a set-up treated in exactly the same way as your experimental ones, but with the independent variable removed. Its job is to prove that the IV caused the effect, and not something else.
Boiling the enzyme is neater than leaving it out altogether: the control tube then contains exactly the same substances, just with the enzyme denatured.
Two more examples of the same idea:
Antibiotics on bacteria. Include a disc soaked in sterile water with no antibiotic. If the bacteria grow normally around it, any clear zone elsewhere must be down to the antibiotic.
Amylase and starch. Include a tube of starch solution with no enzyme. If the starch does not break down on its own, the breakdown you measured elsewhere was caused by the amylase.
Worked examples
WORKED EXAMPLE 1
For an investigation into the effect of light intensity on the rate of photosynthesis in pondweed, state three controlled variables, how you would control each, and why it matters.
Temperature of the waterControl: stand the beaker in a water bath at 25 °C, and place a heat shield between lamp and beaker
Why: the lamp warms the water, and photosynthesis depends on enzymes, so a closer lamp would raise the rate through temperature as well as light
Carbon dioxide availabilityControl: use the same volume of 1% sodium hydrogencarbonate solution, freshly made, for every trial
Why: carbon dioxide is a substrate for photosynthesis and can become the limiting factor instead of light
The pondweed itselfControl: use 5.0 cm lengths cut from the same plant, and leave each for 5 minutes to settle before counting
Why: different pieces have different numbers of chloroplasts, and freshly cut stems release trapped gas that is not from photosynthesis
Each answer names the variable, the method, and the biological reasonThat heat shield point is the one that earns credit — it shows you spotted that the IV and a controlled variable are linked.
WORKED EXAMPLE 2
A student measures pigment leakage from beetroot discs and finds that even the coolest sample gives a strong red colour. Suggest the systematic error and how to remove it.
Step 1: what does “even the coolest” tell you?Every reading is too high by roughly the same amountA shift affecting all readings equally points to systematic error, not random error.Step 2: find the cause
Cutting the discs ruptures cells at the surface, and that pigment washes off into the water regardless of temperature
Step 3: remove itRinse the cut discs in distilled water and blot them before useStep 4: say what will not helpMore repeats — they would simply confirm the same inflated values
Rinse the discs; repeats cannot fix a systematic error
💡 Exam tip
Say how, not just what. “Temperature controlled using a water bath at 30 °C” beats “temperature kept the same”.
Prioritise. Explain the two or three variables that would most affect your DV rather than listing ten.
Include a control run wherever one is possible, and say what it proves.
Mention acclimatising samples before starting — it is a detail few students include.
If you cannot control it, monitor it and record the values.
Use more replicates than you would in chemistry or physics, and say that biological variation is the reason.
⚠ Common mix-up
Thinking repeats fix everything. They reduce random error only; systematic error survives untouched.
Confusing a controlled variable with a control run. One is a factor held constant; the other is a whole extra set-up without the IV.
Forgetting to recalibrate a graticule after changing magnification.
Choosing quadrat positions that look interesting. That is bias — use random coordinates.
Assuming a water bath is instant. Samples need time to reach the set temperature before you begin.
Leaving the control out because “nothing happens in it”. Nothing happening is the result you need.
That completes Stage 1. Up next: Stage 2 — Collecting & Processing Data, where careful design turns into numbers you can actually analyse.
Want this explained one-to-one?
Book a free session with an experienced IB Biology tutor and get your trickiest topics made simple.
