Experimental questions, variables and directional hypotheses

1. Turn a biological observation into a testable question

A useful inquiry starts with an observable biological difference, not simply a topic. “Light and plants” names a topic. “How does daily light duration affect the mean dry mass of radish seedlings after twelve days?” identifies a relationship that can be investigated. The second question specifies an organism, an explanatory variable, a measurable outcome and a time frame. These choices determine what evidence the investigation must collect. A question about dry mass cannot be answered adequately by recording only leaf colour, even when colour changes are interesting.

Begin with a mechanism that could connect the variables. Light supplies energy for photosynthesis, but longer illumination will not necessarily increase growth indefinitely: water, mineral availability, temperature and the plant's response to darkness may also matter. A sensible question tests a range rather than assuming an unlimited relationship. For a classroom investigation, four light durations of 4, 8, 12 and 16 hours might be feasible. Keep the total growth period equal for every group so that age is not a competing explanation.

Distinguish a controlled experiment from an observational comparison. If a researcher assigns light durations, the researcher manipulates the explanatory factor. Comparing plants already growing in sunny and shaded locations is observational: soil, watering and species may differ too. Both approaches can produce useful evidence, but the second supports weaker causal conclusions unless competing explanations are addressed. In Criterion 3, name the exact question and explain why the proposed observations would answer it. A polished method does not rescue a mismatch between the biological question and the evidence collected.

2. Identify the variable and its operational measurement

The independent variable is the factor deliberately changed between conditions. The dependent variable is the outcome used to assess the response. Operationalising a variable means explaining precisely how it is changed or measured. In the seedling example, daily light duration is measured in hours per day, while growth is represented by dry mass in grams after twelve days. “Growth” alone is too broad: length, fresh mass and dry mass describe different aspects of a plant and can lead to different conclusions.

Fresh mass is affected by water content. If treatment groups differ in hydration, a higher fresh mass may not indicate more accumulated organic material. Dry mass reduces that particular ambiguity, although it destroys the sample and therefore prevents repeated mass measurements on the same living plant. Seedling height permits repeated measurements, but a shaded seedling may elongate without gaining proportionately more biomass. Choose the measurement that matches the question, then explain its limitations rather than calling any convenient measurement automatically valid.

Sometimes a biological property is inferred from a proxy. Oxygen volume produced per minute can provide evidence about catalase activity under specified conditions. It does not directly count enzyme molecules. A good response distinguishes the property under investigation, the recorded quantity and the inference joining them. Write units, timing and apparatus into the operational definition: “collect oxygen for sixty seconds in a gas syringe” is more reproducible than “measure bubbles”. If different observers estimate bubble size differently, the measurement method introduces uncertainty even though the independent variable was identified correctly. The current Biology information sheet supports variable identification, but the application must fit the supplied experiment.

3. Write a directional, mechanistically justified hypothesis

A hypothesis is a testable statement about an expected relationship. It is not the investigation question and not a report of results already collected. For the proposed seedling study, a suitable hypothesis is: “Increasing daily illumination from four to twelve hours will increase mean radish seedling dry mass after twelve days because more time is available for photosynthesis, provided other growth requirements remain sufficient.” The direction is explicit, the independent and dependent variables are linked, and the scope is bounded. A prediction is the expected observable pattern if that hypothesis is supported.

The qualification about other requirements is scientifically useful, but it must not become an escape clause that makes the claim impossible to test. Set up sufficient watering and a common nutrient supply before collecting results. State the range for which the directional expectation is reasonable. If the experiment includes sixteen hours, predict a possible plateau separately and justify it using a limiting factor. Do not quietly replace the original hypothesis after seeing an unexpected result. Record the original claim, compare it with the evidence and suggest a revised explanation as a subsequent step.

TASC's external assessment includes hypothesis formulation within Criterion 3. A response such as “light will affect growth” leaves the direction unspecified; “the seedlings will be taller” fails to identify which change produces that outcome. Practise checking three components: what changes, what response is expected, and which direction the response takes. In an unfamiliar enzyme experiment, transfer the structure rather than memorising the plant wording. For example, predict lower initial reaction rate as inhibitor concentration rises, then connect that prediction to reduced successful substrate binding. The biological mechanism makes the claim defensible without turning the hypothesis into an essay.

4. Control competing explanations with specific decisions

A controlled variable is a factor kept sufficiently consistent so that it cannot plausibly explain the treatment difference. In the light-duration study, use one seed variety and batch, comparable initial seedlings, identical growing medium, equal pot volume and a common temperature range. Keep the quantity and timing of water appropriate and consistent. The reason for each decision matters: a temperature difference could alter enzyme-controlled metabolic rates, while unequal water availability could affect stomatal opening and photosynthesis. “To make it a fair test” names an intention but does not explain the biological confounder.

A control group is different from a controlled variable. A control group provides a reference condition against which a treatment is compared. It might receive the ordinary light duration used for that crop; it does not have to receive zero light. A zero-light group can answer a particular question but may also impose extreme stress and therefore be a poor representation of normal conditions. State what comparison the control enables. Keep handling, containers and measurements consistent between the control and treatment groups so that the comparison isolates the intended difference.

Random assignment reduces the chance that initially larger seedlings all enter one group. Randomising pot positions, or rotating them using a predetermined schedule, reduces position effects from uneven illumination or ventilation. Rotation must preserve each group's intended light exposure. Measure room or chamber temperature rather than simply assuming lamps have no heating effect. If a factor cannot be held constant, record it and discuss how it may influence interpretation. A useful evaluation names the factor, predicts the direction of its effect where possible, and proposes a practicable improvement. Adding more seedlings cannot by itself remove a systematic difference in temperature between treatment chambers.

5. Choose replication and sampling that match the comparison

Replication provides several observations within each treatment so that variation between biological specimens becomes visible. Suppose each of four light durations contains eight seedlings. Recording all eight dry masses allows the researcher to calculate a mean and inspect the spread. One plant per treatment cannot distinguish a light effect from the unusual growth of an individual seedling. However, many measured organisms do not necessarily mean many independent experimental units. If all eight plants share one chamber, a chamber-specific temperature fault affects the whole group.

An experimental unit is the smallest unit independently assigned a treatment. When the treatment is applied to an entire chamber, the chamber is the experimental unit; plants within it are subsamples. Several independently managed chambers per light condition provide a stronger comparison than many plants in a single chamber. At school scale, equipment may limit this design. Acknowledge that limitation and avoid claiming independent replication that was not achieved. Repeating the investigation on a later date can reveal whether the pattern persists, although a repeated run also introduces possible seasonal or equipment differences that should be recorded.

Sampling decisions matter before treatment begins. Excluding weak seedlings only after discovering which condition they occupy risks bias. Specify eligibility in advance, such as a defined germination stage without visible damage, then randomly allocate eligible seedlings. Record losses rather than silently replacing inconvenient results. Calculate a group mean from the actual number of measured specimens and report that number. If one group has fewer surviving plants, survival itself may be relevant to the treatment effect. A convincing Criterion 3 answer links the sampling decision to the particular inference: replication estimates biological variability, whereas random assignment reduces systematic initial differences between groups.

6. Build a reproducible method and an honest conclusion

A reproducible method identifies materials, quantities, treatment levels, sequence, timing and measurements clearly enough for another investigator to implement the design. Write the procedure in a logical order: prepare comparable seedlings; allocate treatments; maintain specified conditions; record observations at fixed times; collect the final measurement; calculate the same summary statistic for each group. Include how instruments are checked and how specimen identity is maintained. A label such as “group B” is insufficient unless the method also states which treatment B receives.

Plan the analysis before seeing the data. For daily light duration, a graph with hours on the horizontal axis and mean dry mass on the vertical axis displays the proposed relationship. Record the individual values as well as the means. If the result rises from four to twelve hours but changes little from twelve to sixteen, the conclusion should describe that pattern over the tested range. It should not claim that light always increases growth or that the experiment has proved a universal law. An anomalous value should trigger checking of records and methods, not automatic deletion.

A strong conclusion answers the original question, refers to quantitative evidence, explains whether the hypothesis was supported and identifies an important limitation. For example, quote the difference between group means while noting that one chamber per treatment limits separation of chamber and light effects. A follow-up might repeat the comparison with independently controlled chambers or test whether carbon dioxide becomes limiting. Keep these proposals distinct from changes actually made. The external Biology examination assesses inquiry in biological contexts and excludes the laboratory-safety element of Criterion 3 E1; safe practice remains essential in real school investigations even when that element is not examined.