TL;DR "Human error" is not an acceptable source of error in O-Level Biology Paper 3. Examiners want a specific physical or biological cause. This bank gives you 4 to 6 named errors for each major experiment, the reason the error affects your result, a model ACE sentence, and a concrete improvement suggestion. Use it as a ctrl+F reference: find the experiment, copy the structure, adapt the numbers from your own data.
This page is the per-experiment reference bank: it gives you named errors and model ACE sentences for each Biology experiment. If you are not yet confident about the underlying technique -- what makes an answer valid, the random-versus-systematic taxonomy, five reusable error templates, and how to write a one-sentence mitigation -- read How to Write a Source of Error in a Singapore Science Practical first, then return here for the Biology-specific entries.
Answer first: possible Biology sources of error
Good Biology sources of error name the exact part of the experiment that changed the result. Useful examples include unequal potato cylinder sizes in osmosis, evaporation from uncovered solutions, temperature changes during enzyme work, inconsistent light distance in photosynthesis, air leaks in a potometer, and cross-contamination during food tests.
Do not stop at naming the error. For Paper 3 ACE marks, write one sentence that explains the direction of the effect, then one improvement that would reduce it in a repeat run.
Search-to-answer checkpoint
Use this table when a source-of-error question names the experiment but not the mark scheme language.
Pick the row that matches the actual setup, then add the direction of the effect.
Search or script phrase
Specific source of error
Direction sentence to add
Sources of error in potato osmosis experiment
Potato cylinders had unequal lengths or diameters.
Reviewed by
Ezekiel Tan·Academic Advisor (Biology)
Practical course completion-record note
For practical, lab, and experiment courses, Eclat Institute maintains centre-held attendance records and may also issue an internal attendance or completion document based on participation and internal assessment.
For SEAB private-candidate declarations, the key evidence is the centre's attendance or completion record, not a government-issued certificate.
This is an internal centre-issued certificate, not an MOE/SEAB qualification or accreditation.
Recognition (if any) is determined by the receiving school, institution, or employer.
For SEAB private candidates taking science practical papers, SEAB states you should either have taken the subject before or attend a practical course and complete it before the practical paper date.
The 2026 6093 syllabus puts source-of-error writing inside ACE: students must analyse observations, draw conclusions from data or underlying principles, and evaluate the design. That is why a useful answer cannot stop at "human error" or "more accurate apparatus".
ACE part
What to include
Weak answer to replace
Analyse
Name the measured quantity affected, such as final mass, gas volume, or colour intensity.
"The result is wrong."
Conclude
Link the direction of the error to the Biology process.
"It affects osmosis."
Evaluate
Say whether the result is overestimated, underestimated, scattered, or unreliable.
"There was error."
Refine
Propose a procedure or apparatus change that targets the named cause.
"Repeat the experiment."
This is also how to handle very broad searches like "errors in biology experiments": first identify the experiment family, then choose the specific cause that matches the data pattern.
Paper 3 tuition routing checkpoint
Use this source-of-error bank after marking a practical script. The key question is not whether the student
can list more errors, but whether the student can choose the error that matches the actual data.
Marked-script pattern
What to practise next
When to get structured help
The answer says "human error" or "parallax" for every experiment
Experiment-specific cause selection
The student cannot name what changed in the actual setup.
The error is named but the direction is missing
Effect-on-result sentence
The answer does not say overestimate, underestimate, false positive, or false negative.
The improvement is "repeat the experiment" only
Procedure-specific mitigation
The student cannot name an apparatus or method change.
ACE marks drop across osmosis, enzymes, and potometer work
Cross-experiment error sorting
The same writing weakness appears across more than one practical family.
For Sec 3-4 students who need ACE writing linked to Paper 2 phrasing and Paper 3 technique, use
O-Level Biology tuition Singapore as the main
programme route.
The one rule that earns and loses marks
Before the experiment banks, one rule is worth stating once and stating clearly.
"Human error" is not an acceptable source of error in O-Level Biology Paper 3.
Marking schemes across multiple years consistently reject "human error" as a source of error because it is not specific enough to be actionable. The examiner cannot tell whether you mean parallax on a ruler, inconsistent timing, contamination of a test tube, or something else entirely. Every source of error you write must:
Name a physical or biological cause (e.g. evaporation from the test tube, temperature fluctuation, unequal cylinder lengths).
Explain the direction of the error (e.g. "this causes the final mass to be lower than expected" or "this underestimates the rate of photosynthesis").
Link to a realistic improvement (e.g. "cover the beaker with cling film to reduce evaporation").
The ACE strand of the 2026 SEAB syllabus (6093) explicitly expects "evidence-based conclusions" and "justified refinements." A vague cause earns neither.
How to use this bank
Each experiment section follows the same structure:
Error - the specific physical or biological source of inaccuracy
Why it matters - how the error skews your data, and in which direction
ACE sentence - a model phrase you can adapt for your own written answer
Improvement - a concrete change to procedure or apparatus that reduces the error
For each experiment, you will also find a link to the companion experiment post, which covers method and MMO in full.
Error 1.1 - Unequal potato cylinder lengths or cross-sections
Why it matters: If the cylinders are cut to different lengths or with different diameters, the surface-area-to-volume ratio varies between samples. Cylinders with more surface area relative to volume will exchange water faster. This introduces a random error that scatters your percentage change values and makes the graph curve less smooth.
ACE sentence: "A source of error was variation in the dimensions of the potato cylinders. Cylinders cut to slightly different lengths had different surface-area-to-volume ratios, so some samples gained or lost water faster than others, increasing scatter around the best-fit curve."
Improvement: Use a cork borer of uniform diameter for all cylinders and measure every cylinder with a ruler to the nearest 0.1 cm. Reject any cylinder that differs by more than 0.1 cm from the target length.
Error 1.2 - Surface moisture not dried before weighing
Why it matters: Potato cylinders removed from solution retain a film of liquid on their surface. If this film is not blotted off consistently, it is weighed as if it were part of the cylinder. This makes the final mass appear higher than it should, overestimating mass gain (or underestimating mass loss).
ACE sentence: "A source of error was inconsistent blotting of the potato cylinders before the final weighing. Residual surface moisture added mass that did not reflect actual water uptake by osmosis, causing the calculated percentage change in mass to be higher than expected in some replicates."
Improvement: Blot every cylinder with the same number of strokes of dry filter paper immediately before weighing. Use the same operator for all blotting to keep technique consistent.
Error 1.3 - Evaporation from solution during the soaking period
Why it matters: If the beakers are left uncovered, water evaporates over the soaking period (typically 30 minutes or longer). This raises the concentration of the sucrose solution, meaning the osmotic gradient changes during the experiment. The potato cylinder responds to a concentration that is slightly higher than the labelled value.
ACE sentence: "A source of error was evaporation of water from the uncovered sucrose solutions. As the solvent evaporated, the concentration of each solution increased slightly above the prepared value, causing the potato cylinders to lose more water than expected at lower concentrations and compressing the gradient of the graph."
Improvement: Cover each beaker or test tube with cling film or a stopper during the soaking period to prevent evaporation.
Error 1.4 - Potato tissue not uniform across the same organ
Why it matters: Potato tissue near the skin differs in cell size, vacuole maturity, and starch content compared with tissue from the centre. If cylinders are taken from different zones of the potato, their initial water potential varies, making comparisons between cylinders unreliable.
ACE sentence: "A source of error was variation in the zone of the potato from which each cylinder was cut. Tissue near the surface has a different water potential from core tissue, introducing a systematic variation that cannot be corrected after the experiment."
Improvement: Cut all cylinders from the same depth position in the same potato, working inward from one end only. Discard the outermost 5 mm of tissue before cutting cylinders.
Error 1.5 - Visking tubing not fully wetted before use (osmosis extension)
Why it matters: When Visking tubing (dialysis tubing) is used to model a cell, any section that is stiff or dry has reduced permeability until it is fully wetted. A partially wetted tube restricts water movement at the start of the experiment, making the rate of osmosis appear lower than it actually is.
ACE sentence: "A source of error when using Visking tubing was that tubing stored dry required a soaking period before reaching full permeability. In the early minutes of the experiment, the rate of water movement was lower than expected, causing results at short time points to underestimate the equilibrium rate."
Improvement: Soak Visking tubing in distilled water for at least ten minutes before filling it with the sucrose solution. Check that the tubing is flexible and fully transparent before use.
Error 2.1 - Insufficient heating time in the Benedict's test
Why it matters: The reduction of copper(II) ions in Benedict's solution requires sustained heat at 75 to 95 degrees Celsius. If the water bath is too cool or the heating time is too short, the colour change does not complete. A faint green result might be recorded where the true result is brick-red, leading to an underestimate of the reducing sugar concentration.
ACE sentence: "A source of error was insufficient heating during the Benedict's test. Because the water bath was at approximately 60 degrees Celsius rather than 80 to 95 degrees Celsius, the reaction between copper(II) ions and the reducing sugar was incomplete, producing a green colour that underrepresented the actual sugar concentration in the sample."
Improvement: Check the water bath temperature with a thermometer and allow it to reach a minimum of 75 degrees Celsius before placing the test tubes. Heat for the full two to three minutes specified in the procedure.
Error 2.2 - Contaminated test tubes giving false positive results
Why it matters: If test tubes are not thoroughly rinsed between tests, residue from a previous positive result (for example, a trace of glucose from a Benedict's test) can contaminate the next sample. A false positive means recording a colour change in a tube that should show a negative result.
ACE sentence: "A source of error was contamination from inadequate rinsing of the test tubes between samples. Residual reducing sugar from the previous Benedict's test produced a faint positive result in a sample that should have been negative, leading to an incorrect identification."
Improvement: Rinse each test tube with distilled water three times and allow it to drain completely before reuse. For greater reliability, use a fresh test tube for every sample.
Error 2.3 - Adding biuret reagents in the wrong order
Why it matters: If copper(II) sulfate (CuSO4) is added before sodium hydroxide (NaOH), the copper reacts with hydroxide ions to form a blue copper hydroxide precipitate. This precipitate masks any lilac colour from the protein-copper complex, leading to a false negative for protein.
ACE sentence: "A source of error was adding the biuret reagents in the incorrect sequence. When CuSO4 was added before NaOH, a blue copper hydroxide precipitate formed and obscured the lilac colour that indicates protein, resulting in a false negative result."
Improvement: Always add NaOH first, then add CuSO4 drop by drop while swirling. Standardise the order by labelling the reagent bottles A (NaOH) and B (CuSO4) and always adding A before B.
Error 2.4 - Wet test tube in the emulsion test
Why it matters: The emulsion test relies on fat dissolving in ethanol and forming a cloudy emulsion when poured into water. If the test tube already contains water before ethanol is added, a premature emulsion forms before the deliberate water addition step. It becomes impossible to distinguish a true positive from this procedural artefact.
ACE sentence: "A source of error was water present in the test tube before adding ethanol. The premature mixing of ethanol and water caused a cloudy appearance that was indistinguishable from a genuine fat emulsion, making the positive result unreliable."
Improvement: Dry every test tube used in the emulsion test with a clean paper towel before use, or use test tubes straight from an oven or desiccator.
Error 2.5 - Excess CuSO4 in the biuret test masking the lilac colour
Why it matters: Adding too much copper sulfate keeps the solution a strong blue colour. The lilac colour of a positive protein result is subtle; an excess of blue CuSO4 swamps the lilac and makes the tube appear blue, giving a false negative.
ACE sentence: "A source of error was an excess of CuSO4 added during the biuret test. The intense blue of the excess copper sulfate overwhelmed the lilac colour that indicates protein, causing the result to be recorded as negative when protein was actually present."
Improvement: Add CuSO4 one drop at a time and stop as soon as the colour begins to change. The protocol specifies "a few drops," which in practice means two to four drops. Avoid free-pouring from the bottle.
Experiment 3: Enzyme activity (effect of pH or temperature)
Error 3.1 - Temperature fluctuation during the experiment
Why it matters: Enzyme activity is extremely sensitive to temperature. If the water bath temperature drifts by even 2 to 3 degrees Celsius during the experiment, the rate of reaction changes between readings. This introduces a systematic error that makes it difficult to attribute changes in rate to the independent variable (pH or substrate concentration) rather than temperature change.
ACE sentence: "A source of error was fluctuation in the water bath temperature during the experiment. A drift of approximately 2 degrees Celsius between the first and last readings caused the enzyme activity to change independently of the pH variable being investigated, reducing the reliability of the rate measurements."
Improvement: Use a thermostatically controlled water bath and check the thermometer reading at every time point. Record the actual temperature alongside each reading so any drift can be identified and noted in the evaluation.
Error 3.2 - Enzyme and substrate not pre-equilibrated to the same temperature
Why it matters: If the enzyme solution is at room temperature and the substrate is at the target temperature (or vice versa), the mixture starts at a temperature between the two values. The first readings are taken during the equilibration period, not at the intended experimental temperature. This systematically underestimates or overestimates the rate at the early time points.
ACE sentence: "A source of error was that the enzyme and substrate solutions were not pre-equilibrated to the same temperature before mixing. The reaction began at a temperature approximately 4 degrees below the target, causing the initial rate to be lower than the true rate at the intended temperature."
Improvement: Place both the enzyme and substrate solutions in the water bath for at least five minutes before mixing. Measure the temperature of both solutions immediately before combining them.
Error 3.3 - pH drift during the experiment due to CO2 production
Why it matters: Many enzyme experiments use a carbon dioxide-sensitive indicator or operate in a weakly buffered solution. As CO2 is produced (for example in yeast respiration), it dissolves to form carbonic acid, gradually lowering the pH. The enzyme is therefore acting in a changing pH environment, and the rate you measure reflects an average across a range of pH values rather than the target pH.
ACE sentence: "A source of error was a gradual decrease in pH during the experiment caused by CO2 dissolving in the reaction mixture. This shifted the pH approximately 0.3 units below the intended value over 10 minutes, meaning the measured rate reflected enzyme activity across a range of pH conditions rather than a single controlled pH."
Improvement: Use a buffered solution at the target pH for all enzyme experiments. A phosphate buffer or citrate-phosphate buffer maintains a stable pH even when acidic products are released.
Error 3.4 - Unequal substrate concentrations across replicates
Why it matters: If the substrate solution is not thoroughly mixed before each aliquot is taken, the concentration in each tube differs. Tubes taken from a settled suspension will have a lower concentration near the top. This introduces a random error across replicates and widens the scatter on the rate curve.
ACE sentence: "A source of error was settling of the substrate solution before each aliquot was dispensed. Tubes pipetted from the bottom of the stock received a higher concentration than those taken from the top, causing random variation in the rate measurements that is not attributable to the independent variable."
Improvement: Swirl or gently invert the substrate stock solution immediately before dispensing each aliquot. Use a volumetric pipette rather than a dropper to ensure equal volumes.
Error 3.5 - Timing error at the start of the reaction
Why it matters: In enzyme experiments where the clock is started at the moment of mixing, a delay of even a few seconds between adding the enzyme and starting the stopwatch changes the measured time for the reaction. This makes the calculated rate (e.g. 1/time) slightly too high if the clock starts late.
ACE sentence: "A source of error was a delay between adding the enzyme and starting the stopwatch. A consistent 2 to 3 second delay caused the recorded time to be shorter than the true reaction time, systematically overestimating the rate by approximately 5 to 10 percent in the fastest reactions."
Improvement: Have one person add the enzyme and a second person start the stopwatch simultaneously. Alternatively, use a datalogger that begins recording the moment the enzyme is introduced.
Experiment 4: Photosynthesis rate (DCPIP decolourisation and bubble counting)
Error 4.1 - Unequal leaf disc sizes in the DCPIP assay
Why it matters: If the leaf discs cut for a DCPIP experiment are not uniform in size, they contain different amounts of chlorophyll and therefore have different photosynthetic capacity. Larger discs produce more oxygen or reduce more DCPIP per unit time. This creates a random error across wells and makes rate comparisons between conditions unreliable.
ACE sentence: "A source of error was variation in the area of the leaf discs used in the DCPIP assay. Discs cut with a slightly larger diameter contained more chlorophyll and produced oxygen faster, causing the rate of DCPIP decolourisation to be overestimated in those wells compared with the standard condition."
Improvement: Use a standard-sized cork borer (for example 1 cm diameter) for every disc. Avoid using visibly damaged or discoloured sections of the leaf.
Error 4.2 - Inconsistent counting of bubbles in pondweed experiments
Why it matters: In a bubble counting experiment using Elodea or similar pondweed, bubbles of oxygen emerge at irregular intervals. Counting bubbles by eye over a fixed period introduces a random error: a bubble that forms at the very end of the counting interval may be counted or missed depending on the observer. Different observers may also count partial bubbles differently.
ACE sentence: "A source of error was inconsistent bubble counting near the end of the measurement interval. Bubbles that appeared in the final second of the count were included by some observers and missed by others, producing a random variation of approximately 2 to 3 bubbles per minute between replicates."
Improvement: Use a datalogger with an oxygen sensor probe instead of counting bubbles by eye. If bubble counting is required, standardise counting by ruling that only bubbles that leave the stem completely within the timed interval are counted, and use a single observer for all readings.
Error 4.3 - Background respiration of the pondweed reducing the apparent photosynthesis rate
Why it matters: The rate of oxygen production measured by bubble counting or DCPIP decolourisation is the net rate of photosynthesis (gross photosynthesis minus respiration). The plant is respiring continuously, consuming some of the oxygen it produces. At low light intensities, the net rate may be close to zero or even negative. If the experiment does not account for this, the measured rate underestimates gross photosynthesis.
ACE sentence: "A source of error was that the measured rate of oxygen production reflected the net rate of photosynthesis rather than the gross rate. At the lowest light intensity, the plant consumed oxygen through respiration at approximately the same rate as it was produced, causing the bubble count to approach zero and masking any further decrease in gross photosynthesis."
Improvement: To measure gross photosynthesis, also measure the rate of respiration in the dark for the same sample and add it to the net rate. Acknowledge in the ACE evaluation that all measurements are net values.
Error 4.4 - Light source not at a consistent distance from the pondweed
Why it matters: Light intensity follows an inverse square relationship with distance. Moving the lamp just 2 cm closer at the lower distance settings (e.g. 10 cm) changes the intensity by a large proportion. If the distance is not measured from the lamp filament to the same reference point on the plant every time, the effective light intensity at the leaf surface varies.
ACE sentence: "A source of error was inconsistency in measuring the distance between the lamp and the pondweed. Because the distance was measured to the front of the lamp housing rather than to the filament, the effective light intensity at the leaf surface differed from the calculated value, causing the rate to be overestimated at shorter distances."
Improvement: Fix the lamp in a retort stand and measure the distance to the same reference point on the plant (e.g. the tip of the cut stem) using a ruler clamped horizontally. Keep the plant stationary during readings.
Error 4.5 - Heat from the lamp raising the water temperature
Why it matters: As the lamp distance decreases, it also increases the temperature of the water around the pondweed. Higher temperatures increase enzyme activity (up to the optimum) and increase the solubility change in CO2. This means that apparent increases in photosynthesis rate at higher light intensities may partly reflect a temperature effect rather than a light effect.
ACE sentence: "A source of error was that placing the lamp closer to the pondweed also increased the temperature of the surrounding water. An increase of approximately 3 degrees Celsius at the closest setting increased both the rate of photosynthesis and the rate of respiration, making it impossible to attribute the change in bubble count solely to the change in light intensity."
Improvement: Place a heat filter (a flat-sided container of water) between the lamp and the plant to absorb infrared radiation. Monitor the water temperature with a thermometer at every setting and record it alongside each rate measurement.
Error 5.1 - Air bubbles introduced during assembly
Why it matters: The potometer measures water uptake by timing the movement of an air-water meniscus along a capillary tube. If air bubbles are introduced during assembly (by lifting the shoot out of water even briefly), additional bubbles move alongside the meniscus, making the apparent water uptake appear higher than it is.
ACE sentence: "A source of error was the introduction of air bubbles during assembly of the potometer. An air bubble trapped behind the main meniscus caused the bubble to advance more rapidly than expected, overestimating the rate of water uptake by the shoot."
Improvement: Assemble the potometer entirely underwater. Cut the shoot underwater with a clean, oblique cut and transfer it directly into the potometer without exposing the cut end to air.
Error 5.2 - Stomata not fully open at the start of the experiment
Why it matters: Stomatal aperture changes in response to light and CO2 concentration. If the plant has been in the dark or in low-CO2 conditions before the experiment, the stomata may be partially closed. The measured transpiration rate during the initial readings will be lower than the true rate for the environmental conditions being tested.
ACE sentence: "A source of error was that the stomata of the shoot were not fully open at the start of the experiment. Because the plant had been stored in a dark laboratory overnight, the guard cells had not yet responded to light, causing the initial transpiration rate to underestimate the steady-state rate by approximately 20 to 30 percent."
Improvement: Place the plant under the experimental light conditions for at least 20 minutes before starting measurements to allow the stomata to equilibrate. Begin timing only when the rate of bubble movement is stable.
Error 5.3 - Water loss through the cut stem surface rather than by transpiration
Why it matters: Some water evaporates directly from the cut surface at the base of the shoot, particularly if the cut end is exposed to the air or if the seal between the shoot and the potometer tube is imperfect. This water loss is recorded as if it were transpiration through the leaves, overestimating the transpiration rate.
ACE sentence: "A source of error was water evaporation from the cut end of the shoot. A small gap between the shoot and the rubber connector allowed water to move directly from the cut surface into the surrounding air, inflating the measured rate of water uptake beyond the true transpiration rate."
Improvement: Seal the connection between the shoot and the potometer tube with petroleum jelly (Vaseline). Check visually that no liquid is seeping around the seal before beginning measurements.
Error 5.4 - Not waiting for steady state before recording readings
Why it matters: When the potometer is first assembled, the shoot is adjusting to the new conditions and the meniscus may move irregularly. Taking readings before the rate of movement has stabilised means recording a transitional rate rather than the true steady-state transpiration rate for those conditions.
ACE sentence: "A source of error was taking the first reading before the transpiration rate had reached a steady state. The meniscus moved faster during the first five minutes as the shoot responded to being cut and reassembled, overestimating the true transpiration rate at the experimental condition."
Improvement: Allow five to ten minutes after assembly (or after each change of condition) before starting the timed reading. Reject the first measurement if it differs substantially from the second.
Why it matters: A respirometer measures changes in gas volume (typically the decrease in oxygen as it is consumed by respiring organisms). Any leak in the tubing, stoppers, or syringe connections allows air to enter or escape, changing the gas volume independently of respiration. This makes the recorded volume change smaller than expected, underestimating the rate of respiration.
ACE sentence: "A source of error was a small leak at the junction between the rubber tubing and the manometer tube. Atmospheric air entered the system, partially compensating for the decrease in gas volume caused by oxygen consumption, causing the rate of respiration to be underestimated by approximately 15 percent."
Improvement: Before the experiment, seal all connections with petroleum jelly and check for leaks by briefly blocking the outlet and observing whether the manometer reading drifts. Replace any stopper that does not create an airtight seal.
Error 6.2 - Temperature change affecting gas volume independently of respiration
Why it matters: Gas volume changes with temperature (Charles's Law). If the ambient temperature changes during the experiment (for example due to sunlight through a window or air conditioning cycling), the gas volume in the respirometer tube changes for a physical reason unrelated to respiration. This is a systematic error that mimics or masks the effect of oxygen consumption.
ACE sentence: "A source of error was a rise in room temperature of approximately 1.5 degrees Celsius during the experiment. This caused the gas in the respirometer to expand slightly, partially offsetting the decrease in volume due to oxygen consumption and causing the calculated respiration rate to be underestimated."
Improvement: Set up a thermobarometer (a control tube containing dead organisms of the same mass) alongside the experimental tube. Any volume change in the thermobarometer reflects temperature and pressure changes only. Subtract these readings from the experimental readings.
Error 6.3 - CO2 not absorbed effectively by the soda lime
Why it matters: Respirometry experiments typically include soda lime or potassium hydroxide to absorb CO2 produced during respiration. If the soda lime is old, damp, or insufficient in quantity, some CO2 remains in the gas phase. Because CO2 released by respiration partially replaces the O2 consumed, the net decrease in gas volume is smaller than the volume of O2 consumed alone, underestimating the respiration rate.
ACE sentence: "A source of error was deteriorated soda lime that did not fully absorb the CO2 produced by the germinating seeds. The residual CO2 occupied approximately 0.3 cm3 of the total gas space, causing the measured decrease in gas volume to underestimate the volume of O2 consumed and therefore the rate of aerobic respiration."
Improvement: Use freshly prepared soda lime pellets and ensure they are placed in a separate chamber upstream of the organisms so that CO2 is absorbed before it can mix with the remaining O2. Replace soda lime at the start of every experiment.
Error 6.4 - Variation in mass or developmental stage of organisms used
Why it matters: Respiration rate depends on metabolic activity, which varies with the mass and developmental stage of the organism. If germinating seeds are at different stages of germination (some just starting, some with radicles 5 mm long), their respiration rates differ. Using the same total mass of organisms does not control for this variation if the organisms are not at the same stage.
ACE sentence: "A source of error was variation in the germination stage of the seeds used. Seeds with radicles approximately 10 mm long had a higher metabolic rate than seeds that had just begun to germinate, causing the mean respiration rate to reflect a mixture of developmental stages rather than a single controlled condition."
Improvement: Select seeds that are at the same visible stage of germination (for example, radicle 3 to 5 mm in all cases) and weigh them individually to confirm that the total mass is consistent between replicates.
Quick reference: common error types and their direction
