QCE Biology - Unit 3 - Biodiversity and populations

Sampling Methods and Bias | QCE Biology

Learn random, systematic and stratified sampling, quadrats, transects, capture-recapture and bias control in QCE Biology.

Updated 2026-05-18 - 6 min read

Sampling methods and bias is part of the way QCE Biology turns living systems into evidence students can describe, analyse and evaluate. The safest way to study it is to connect each term to a data pattern, a biological mechanism and a limitation.

Core explanation

Random sampling

Random sampling gives every location a fair chance of selection. It is useful when a habitat is fairly even and the aim is to estimate abundance without the researcher choosing convenient or attractive sites.

Systematic sampling

Systematic sampling collects data at fixed intervals. A line transect across a rocky shore is systematic because the researcher records species at regular points along an environmental gradient.

Stratified sampling

Stratified sampling separates a study area into meaningful zones first, then samples inside each zone. This is stronger when a site contains clear microhabitats such as shaded forest, creek edge and open grass.

Bias control

Bias is reduced by increasing sample number, standardising counting rules, using random-number generators, calibrating equipment and recording the precision of instruments.

Matching method to organism and question

| Method | Best for | Data produced | Main assumption | | --- | --- | --- | --- | | Quadrat count | Sessile or slow-moving organisms | Abundance, density, richness, frequency or cover | Quadrats represent the wider area | | Point transect | Changes along a gradient | Presence or abundance at fixed points | The transect crosses the relevant gradient | | Belt transect | Zonation across a gradient | Continuous bands of quadrat data | Sampling width is consistent | | Capture-recapture | Mobile animals | Population estimate | Marks are retained and do not affect survival or recapture | | Stratified random sampling | Mixed habitats | More representative site estimates | Strata match real ecological differences |

Sampling design should match the spatial scale and temporal scale of the claim. A single afternoon sample cannot confidently describe seasonal abundance. A sample from one creek bank cannot automatically represent the whole catchment. Repeated sampling increases confidence because it shows whether a pattern is consistent or just a temporary fluctuation.

Common sources of bias include choosing convenient sites, avoiding difficult terrain, placing quadrats after seeing organisms, using inconsistent counting rules for partly covered quadrats, sampling only one time of day and disturbing organisms before counting them.

Reliability, validity and precision in fieldwork

Reliability improves when repeated sampling gives similar results. Increasing the number of quadrats, using repeated transects and sampling on multiple days can reduce the effect of chance. Validity improves when the method actually measures the claim being made. For example, percentage cover is valid for estimating plant dominance, but it is not a direct measure of genetic diversity.

Precision is the fineness of the measurement. A pH probe reporting to two decimal places is more precise than universal indicator paper, but that does not automatically make the investigation valid if the wrong sites were sampled. Accuracy depends on whether measurements are close to the true value, which requires calibration and consistent technique.

Fieldwork answers should also mention ethical and practical constraints. Marking animals should minimise harm, handling time and stress. Sampling protected or dangerous species may require indirect evidence such as tracks, camera traps or environmental DNA. In those cases, the method may estimate presence more easily than abundance.

When comparing sites, keep the sampling effort equal unless the method deliberately standardises by area or time. More quadrats at one site can make that site appear richer simply because it was searched more thoroughly.

How to use this in data questions

Start by identifying what has been measured. In Biology, a graph or table is rarely just asking for a trend; it is asking whether you can connect the trend to a process. Quote enough data to show the pattern, then use the concept language from the syllabus. If the evidence is limited, name the limitation precisely: sample size, sampling method, uncontrolled variables, measurement precision, population choice or the time scale of the data.

A useful study habit is to turn each heading into a data prompt. Ask what you would expect to happen if the relevant variable increased, decreased or was removed. For ecology topics, think about abundance, distribution, biodiversity, biomass and carrying capacity. For genetics topics, think about genotype, phenotype, gene expression, allele frequency and inheritance pattern. For evolution topics, think about variation, selection pressure, gene flow, isolation and relatedness.

When a question asks you to evaluate, do not just list problems with the experiment. Link the limitation to the confidence of the conclusion. For example, a small sample size matters because a few unusual individuals can distort the pattern. An uncontrolled abiotic factor matters because it gives another possible explanation for the same biological trend. This is the difference between naming a limitation and using it scientifically.

Worked example

Common exam traps

Other traps to watch for:

• using a general word when a syllabus term is available
• ignoring units, sample size or time scale
• treating a model as a perfect copy of the real ecosystem or cell
• writing a memorised paragraph that does not use the given data

Quick check

Sources

• QCAA Biology subject page
• QCAA Biology 2025 syllabus
• OpenStax Biology 2e: Population Demography
• Biology LibreTexts: Sampling ecosystems/03%3A_Ecology/3.01%3A_Ecosystem_Ecology)
• Australian Museum: Biodiversity