Variables and controls: what an experiment can show

Independent, dependent and control variables

An investigation identifies an independent variable, a dependent variable and the relevant control variables. The independent variable is changed or selected; the dependent variable is measured; control variables are kept constant where they could affect the result. State units for measured quantities where applicable.

Independent variable

The quantity you change on purpose, and the only one that should differ between your treatments. It goes on the horizontal axis.

Dependent variable

The quantity you measure, whose value depends on the independent variable. It goes on the vertical axis.

Control variable

A quantity held the same across every treatment, so that it cannot be the reason the dependent variable changed.

Control experiment

An extra run of the method, differing in one deliberate way, done to rule out an explanation the main runs cannot rule out on their own.

Describe the drawing

An investigation drawn as the graph it produces. Five chosen levels of one variable are marked along the horizontal axis, dashed lines rise from each to a point on a curve that climbs steeply at first and then flattens, and the vertical axis carries the quantity being measured. A panel on the right lists four quantities held constant: temperature, pH, enzyme concentration and total volume.

Choose enough values of the independent variable to reveal the expected relationship, spread across an appropriate range. Five values are often suitable for drawing a trend, but the method and specification take priority. Four points can be joined by a straight line that a fifth would rule out, and the range must cover the region of interest: an enzyme investigation running only from 20 °C to 30 °C gives a rising line and misses an optimum near 40 °C.

Use a bar chart for categorical groups and a scatter graph or line graph for continuous data, as appropriate to the design. Species of plant and brand of antiseptic are categorical; temperature, concentration and time are continuous. A bar chart of a continuous variable does not show the shape of the relationship.

Select control variables that could plausibly affect the dependent variable and state how each is controlled. For example: 'Keep each tube in a 35 °C water bath for five minutes before starting the reaction.' This is reproducible because it gives both the set value and the method.

A control variable is not a control experiment

Controlling variables does not answer a further question: would the result have appeared without the factor being tested? That is the function of a control experiment. The two standard kinds have opposite functions.

A negative control is the run in which the factor you are testing has been removed or destroyed, with everything else left exactly as it was. Boiled enzyme, distilled water in place of the extract, an agar disc soaked in nothing but the solvent. If the effect shows up in the negative control, the effect was never yours: the apparatus, the solvent or the handling produced it. A negative control guards against a false positive.

A positive control uses something already known to produce the effect: a disc of an antibiotic the organism is sensitive to, a glucose solution in a set of Benedict's tests. If it shows nothing, the method failed and the flat readings from your real samples mean nothing at all. A positive control guards against a false negative.

Neither control substitutes for the other. If six plant extracts are tested for antibacterial activity and no zones appear, a blank solvent disc showing nothing is the expected result. But a lawn that never grew, discs that never dried and a cold incubator would produce the same appearance. Only a disc of a known antibiotic distinguishes the two cases.

Confounding: when one change is really two

A confounding variable is one that changes along with your independent variable and could produce the same effect on its own. It is a serious fault in a design, because it does not appear as scatter or as an anomalous result: it produces a consistent graph of a relationship that may not exist.

Describe the drawing

Three boxes joined by arrows. A box for moving a lamp closer sits top left, a box for the bubble count rising sits top right, and a straight arrow runs between them. A third box below records that the water also warms, with arrows running into it from the lamp and out of it to the bubble count, so two separate routes connect the same cause to the same effect.

The pondweed and the lamp is the standard case. Move the lamp closer and the light intensity rises, and so does the temperature of the water, because a filament lamp is mostly a heater. Oxygen production rises. You cannot tell from the readings whether it rose because of the light or because of the warmth, and no number of repeats will separate them, because every repeat changes both things together in exactly the same way. The fix is in the design: put a heat shield of water between the lamp and the beaker, or use an LED source, and then measure the temperature to show it stayed put.

It is also what makes some biological questions genuinely hard. Ask whether people who eat oily fish have fewer heart attacks and you are also asking about people who can afford oily fish, who cook at home, who were told by a doctor to change their diet. Ruling those out means measuring them and comparing people who match on them, which is why an epidemiological paper spends most of its length on adjustment and a paragraph on the result.

Confounding variable

A variable that changes with the independent variable and could itself explain the change in the dependent variable, so the two effects cannot be separated by the data collected.

Randomisation

Assigning subjects or sample positions by chance rather than by choice, so that variables not identified in advance tend to balance out between the groups on average.

Randomisation addresses confounding variables that were not identified in advance. Assign which leaf disc goes into which concentration by drawing lots rather than by picking, and the discs that happen to be thicker, older or nearer the midrib tend to be spread across every treatment instead of piling into one. Note the word 'tend': randomisation balances confounders in expectation, across many runs; it does not guarantee an even spread in any single experiment, which is one reason sample size still matters.

Correlation and causation, argued rather than asserted

A correlation does not by itself show that one variable caused the other. A common cause, reverse causation or chance may explain the association. A causal conclusion becomes stronger when several independent forms of evidence agree, including a plausible mechanism, the correct time order, a dose-response relationship, adjustment for confounding variables, replication in other populations and reversal when the exposure is removed. Where random allocation would be unethical or impossible, causal inference may depend on this combination of evidence rather than on a single experiment. The link between smoking and lung cancer is the standard example: no randomised trial assigned people to smoke, and the causal claim rests on that combination.

Correlation

A relationship in which two variables tend to change together. It may be positive or negative, strong or weak, and it makes no claim about which variable moved first.

Causation

A relationship in which changing one variable changes the other. It cannot be read off a scatter graph.

Three things produce correlations with no causal link between the variables plotted. A common cause may sit behind both, as the summer sits behind ice cream sales and drowning deaths. The arrow may run the other way, which is reverse causation, as when a nutrient is low in people with a disease because the disease stopped them absorbing it. And with enough variables plotted against enough others, some pairs correlate by coincidence, which is why a relationship found by trawling data has to be tested on a fresh set.

Several separate lines of evidence, none decisive alone, strengthen a causal claim:

An A-level answer does not need all six. It needs the specific alternative explanation named, whether common cause, reverse causation or coincidence, together with the evidence that would test it.

Validity, accuracy, precision and the two kinds of repeat

Validity, accuracy, precision, repeatability and reproducibility are five distinct properties. They fail in different ways, so identifying which one an investigation lacks identifies what to change.

Describe the drawing

A two by two grid of target boards, each with three rings and four shots marked on it. The left column has shots clustered tightly together and the right column has them widely scattered; the top row has shots centred on the bullseye and the bottom row has them displaced to one side. All four combinations of the two properties appear.

Valid

A measurement is valid if it measures what it claims to measure; an investigation is valid if its design allows its conclusion to follow. A confounding variable destroys validity however careful the measuring was.

Accurate

Close to the true value.

Precise

Repeat readings close to one another. Precision says nothing at all about being right.

Repeatable

The same person, using the same method and the same apparatus, gets the same results again.

Reproducible

Somebody else, or the same person using different apparatus or a different method, gets the same results.

Resolution (of an instrument)

The smallest change an instrument can register: 0.01 g on a two-decimal-place balance, 1 mm on a metre rule.

The two kinds of error map onto two of those words. A random error scatters readings above and below the true value, as in judging a meniscus, starting a stopclock late, or tissue that was not quite identical. It damages precision, and a mean of several repeats reduces its effect because the scatter partly cancels. A systematic error shifts every reading the same way: a balance that was not zeroed, a thermometer reading 2 °C high, a bubble in a burette tip. It damages accuracy, and repeating the measurement a hundred times gives a hundred readings wrong by the same amount. That is the lower-left case, and it is why precision is not evidence of accuracy.

Reliability is used loosely to mean that repeats agree, and is accepted, but repeatable and reproducible are more precise terms. An anomaly is a reading that does not fit the pattern of the others: identify it, state that it is anomalous, and exclude it from the mean only with a stated reason, such as a bubble in the tube or a sample that was dropped. Excluding a point solely because it does not fit the line is not justified.

What the experiment can actually show

A complete write-up states the limits of the result. A conclusion that extends beyond the data is unsupported however carefully the data were collected.

Your result holds over the range you tested and for the material you used. If your temperatures ran from 20 °C to 50 °C, the data say nothing about 60 °C. Reading a value off an extended line is extrapolation and is not supported by the measurements; reading between two measured points is interpolation and is. Catalase from potato is not catalase from liver, and cultured cells are not tissue, so name the organism and the tissue.

It describes what happened, not why. An investigation showing that rate falls above 45 °C shows that rate falls above 45 °C; denaturation is an explanation brought to it from elsewhere, and the careful sentence says 'this is consistent with the enzyme denaturing' rather than 'this shows the enzyme denatured'.

And it holds for the sample you took. Five leaf discs from one leaf describe that leaf. Extending the claim to the plant or the species needs a sample drawn to represent them, which is a decision made before any measuring starts and cannot be repaired afterwards.

Measuring a living animal: caffeine and the heart rate of Daphnia

Daphnia, the water flea, is a freshwater crustacean a few millimetres long with a transparent body. Its heart can be watched beating through a light microscope with no dissection, which is why Pearson Edexcel Biology A sets the effect of caffeine on heart rate in Daphnia as Core Practical 1. The design follows the earlier sections of this lesson. A living animal adds three problems of its own.

The first problem is heat. The microscope lamp warms a small drop of water quickly, and Daphnia is an ectotherm whose heart rate rises with temperature, so time under the lamp becomes a confounding variable. Keep each count short, switch the lamp off between counts or use a cool LED source, and measure the temperature of the water at the start and end.

The second is that the heart beats several times a second, too fast to count steadily for a full minute. Count for 15 or 20 seconds and scale up, or tap a pencil on paper once per beat and count the dots afterwards. Scaling up multiplies any miscount too, so repeat each count and take the mean.

The third is that two animals in the same solution do not beat at the same rate. Record each animal's rate in pond water first, then in caffeine solution, and report the change from its own baseline. Use a fresh animal for each concentration so that no caffeine is carried over from a previous solution. A few strands of cotton wool in the drop on a cavity slide hold the animal still without crushing it.

Every practical also needs a risk assessment, which separates the hazard from the risk and names a control measure for each. Here the hazards are modest: broken slides and coverslips, mains electricity close to water, pond water carrying microorganisms, and caffeine itself, which is harmful if swallowed as a solid. Wash hands after handling pond water, keep liquids away from the microscope's plug, and check the CLEAPSS Hazcard and your school's risk assessment for the caffeine stock.

Hazard

Anything with the potential to cause harm, such as a chemical, a sharp edge or a pathogen.

Risk

How likely it is that a hazard causes harm, and how serious the harm would be, under the conditions in which the work is done. Control measures reduce the risk; the hazard stays the same.

The ethics of using invertebrates in research

Pearson Edexcel Biology A pairs Core Practical 1 with statement 1.17: discuss the potential ethical issues regarding the use of invertebrates in research. A discussion needs arguments on both sides, each tied to a fact about the animal or the work, and a judgement.

Sentience

The capacity to have experiences that feel good or bad to the animal itself, such as pain, distress or comfort.

The 3Rs

Replacement, Reduction and Refinement: replace animals with other methods where possible, reduce the number used, and refine procedures so that the animals used suffer as little as possible. Set out by Russell and Burch in 1959.

The central question is whether the animal can suffer, and for most invertebrates the answer is not known. The evidence used to judge sentience includes receptors that detect damaging stimuli, a centralised nervous system that integrates information, and behaviour that goes beyond a reflex, such as guarding an injured part or accepting a cost to avoid a harmful stimulus. A Daphnia has a small brain and responds to light and touch, but that evidence has not been gathered for it in the detail it has for crabs, lobsters and octopuses.

The law in the UK draws its line by animal group. The Animals (Scientific Procedures) Act 1986 regulates procedures on living vertebrates and, since 2013, on cephalopods such as octopus and squid. The Animal Welfare (Sentience) Act 2022 treats all vertebrates, cephalopod molluscs and decapod crustaceans, such as crabs and lobsters, as sentient for the purposes of UK government policy. Daphnia is a crustacean but not a decapod, so neither Act covers it: using it in a school laboratory needs no licence. That it is legal does not settle whether it is right.

Two ethical frameworks approach the question differently. A utilitarian argument weighs the benefit, here learning a method or testing a drug cheaply, against the total harm, and accepts the work if the benefit is larger. A rights-based argument asks whether the animal has a right not to be harmed that no benefit can override, and that turns again on whether it can suffer.

Whichever view is taken, the 3Rs give the practical steps. Use the fewest animals that give a reliable result. Keep each one in pond water, out of the lamp's heat and for as short a time as possible. Do not use concentrations far beyond the range the question needs, and return each animal to a separate culture tank to recover rather than to the stock it came from.