Planning, Analysis & Evaluation
Paper 5 (1 hour 15 minutes, 30 marks) tests your ability to plan experiments, analyse data, and critically evaluate procedures — the three pillars of scientific practice. Questions are based on unfamiliar contexts drawn from any topic in the A Level specification. You will not have access to a laboratory, but you will need to apply your knowledge of experimental design, statistical tests, graph interpretation, and sources of error to novel scenarios. Every mark is contestable; clear, precise scientific language wins.
Planning
Paper 5 structure and marks
| Section | Skill tested | Typical marks | What is required |
|---|---|---|---|
| Planning (P) | Experimental design | ~10–12 marks | Hypothesis, variables, method, controls, safety, predicted results |
| Analysis (A) | Data processing and interpretation | ~10–12 marks | Calculations, graph plotting, drawing conclusions from data |
| Evaluation (E) | Critical assessment and improvement | ~8–10 marks | Identify limitations, sources of error, suggest improvements, assess reliability |
- Questions use unfamiliar biological contexts — do not panic; apply general experimental principles to the new scenario
- Marks are earned by specificity: "use a colorimeter set to a wavelength of 650 nm" earns a mark; "measure the colour" does not
- Always link your planning to the biology: explain why a variable matters, not just what it is
- Time management: allocate roughly 15 minutes per section; don't spend all time on Planning at the expense of Analysis and Evaluation
Hypothesis and prediction
A good hypothesis is a testable, specific, directional statement that identifies the independent variable (IV) and dependent variable (DV) and predicts the relationship between them with biological justification:
Structure: "If [IV] increases / decreases, then [DV] will increase / decrease, because [biological reason]."
Weak (0 marks): "Temperature affects enzyme activity."
Strong (2 marks): "If temperature increases from 20°C to 50°C, the rate of enzyme-catalysed hydrolysis will increase, because higher temperature increases the kinetic energy of substrate and enzyme molecules, leading to more frequent effective collisions between substrate and active site."
Prediction (separate from hypothesis): "Above the optimum temperature (~40°C for most mammalian enzymes), the rate will decrease as the enzyme denatures and loses its tertiary structure, causing the active site to change shape so that the substrate can no longer bind."
Variables — identifying and controlling
The variable you deliberately change. Must be precisely defined with units and range: "temperature, varied from 10°C to 60°C in steps of 10°C" not just "temperature".
Aim for at least 5 different values to show a trend; more values provide a more reliable picture of the relationship.
The variable you measure as a result of changing the IV. Must be quantifiable with units and method: "the absorbance of the reaction mixture measured at 540 nm using a colorimeter" not just "how much colour". State how you will measure it.
All other variables that must be kept constant to ensure a fair test. For each controlled variable, state: (1) what it is, (2) why it needs to be controlled, (3) how it will be controlled.
Example: "pH — pH affects enzyme tertiary structure and ionisation of R-groups; control by using a buffer solution of pH 7 throughout."
A replicate in which the IV is absent or set to zero — everything else is identical. Confirms that any change in the DV is due to the IV and not some other factor. For enzyme experiments: a control using boiled enzyme (denatured) or no substrate confirms the reaction requires active enzyme.
Writing the method
- Materials and quantities: state specific concentrations, volumes, masses — not vague amounts. "5 cm³ of 1% starch solution" not "some starch"
- Measurement technique: name the instrument and how it is used. "Use a colorimeter set to 660 nm to measure the absorbance of each sample" not "use a device to check colour"
- Replicates: state that each condition will be repeated at least 3 times and the mean calculated. Explains why: reduces the effect of random variation and increases reliability
- Order of operations: number your steps; they should be in a logical sequence that could actually be followed by another scientist
- Time point / endpoint: state when measurements are taken. "Record absorbance every 30 seconds for 5 minutes" not "take measurements"
- Negative control and blank: for colorimetry, zero the colorimeter with a water/buffer blank; state this explicitly
Reliability, validity and accuracy
| Term | Definition | How to improve |
|---|---|---|
| Reliability | Results are reproducible — similar results are obtained when the experiment is repeated | Repeat measurements (triplicates or more); calculate mean; identify and exclude anomalous results; use precise instruments |
| Validity | The experiment actually tests what it claims to test; the DV change is caused only by the IV | Control all other variables; use appropriate controls; ensure the method measures the intended quantity |
| Accuracy | Results are close to the true value | Calibrate instruments; use appropriate measurement tools; minimise systematic errors |
| Precision | Results are consistently close to each other (small spread) | Use more precise instruments; standardise technique; reduce random errors |
Risk assessment
Paper 5 may ask for a risk assessment. For each hazard, identify the harm, the risk level, and the precaution:
| Hazard | Harm | Precaution |
|---|---|---|
| Hot water bath at 60°C | Scalds to skin | Use tongs to handle test tubes; do not lean over the water bath; keep water level below the rim |
| Iodine solution | Stains skin and clothing; mild irritant | Wear gloves and lab coat; rinse skin immediately with water if contact occurs |
| Concentrated acid (e.g. HCl) | Corrosive; burns to skin and eyes | Wear safety goggles and gloves; work in fume cupboard; have access to eyewash station |
Design an experiment to investigate the effect of substrate concentration on the initial rate of an enzyme-catalysed reaction, using hydrogen peroxide (H₂O₂) as the substrate and catalase enzyme (from potato tissue).
Your answer should include: (a) hypothesis [2], (b) identification of IV, DV, and two controlled variables [4], (c) one safety precaution [1], (d) how you would ensure reliability [1].
(a) Hypothesis [2 marks]
- As the concentration of hydrogen peroxide increases, the initial rate of the catalase reaction will increase, because more substrate molecules are present; more enzyme-substrate complexes can form per unit time (more frequent successful collisions between substrate and active site) [2]
- Above a saturating concentration, the rate will plateau as all active sites are occupied and enzyme (not substrate) is the limiting factor [accept as bonus]
(b) Variables [4 marks; 1 each]
- IV: concentration of H₂O₂ solution, e.g. varied from 0.1% to 2.0% in steps of 0.3% using serial dilution with water [1]
- DV: initial rate of O₂ production; measured as the volume of O₂ collected (cm³) in the first 30 seconds using an inverted measuring cylinder or gas syringe [1]
- Controlled variable 1: mass/volume of potato tissue used — controls enzyme concentration; cut equal discs of potato tissue (e.g. 5 discs of 5 mm diameter) for each replicate [1]
- Controlled variable 2: temperature — temperature affects enzyme activity; maintain constant temperature using a water bath set to 25°C [1]
(c) Safety precaution [1 mark]
- H₂O₂ is an oxidising irritant — wear safety goggles and gloves; avoid skin contact; have eyewash station available [1]
(d) Reliability [1 mark]
- Repeat each concentration at least 3 times and calculate the mean volume of O₂ produced; discard any anomalous values and repeat to reduce the effect of random variation [1]
Analysis & statistics
Data tables
- Column headings: include the variable name AND unit in the heading (not in the data cells): "Temperature / °C" not "Temperature" or "Temperature (°C)"
- IV in the left column(s); DV and calculated results in subsequent columns
- Consistent decimal places: all values in a column to the same number of significant figures / decimal places, consistent with the precision of the measuring instrument
- Mean column: include a column for the mean of replicates; show the mean to the same number of decimal places as the raw data
- No units in data cells: units belong only in the column heading
- Raw data AND processed data: include raw readings and any calculated values (e.g. rate = 1/time) in separate columns, clearly labelled
Graphs and plotting
| Situation | Graph type | Reason |
|---|---|---|
| Continuous IV and DV (e.g. temperature vs reaction rate) | Line graph (scatter plot with line of best fit) | Shows trends and relationships between continuous variables |
| Discontinuous / categorical IV (e.g. different species, named treatments) | Bar chart | Discrete categories cannot be placed on a continuous scale |
| Showing proportions of a whole | Pie chart | Used rarely in biology practicals; only when parts of a whole are compared |
| Frequency distribution of a continuous variable | Histogram | No gaps between bars; shows distribution of one continuous variable |
- Axes: IV on x-axis; DV on y-axis. Label both with variable name AND units: "Temperature / °C" and "Rate of reaction / cm³ min⁻¹"
- Scale: use at least half the grid; choose a scale that is easy to read (multiples of 1, 2, 5, or 10 per grid square); start at 0 unless data dictates otherwise
- Plotting: plot each point accurately; use small crosses (×) or dots (•) — not large blobs; if plotting mean ± SD or range, add error bars
- Line of best fit: draw a smooth curve or straight line that best fits the trend, not a dot-to-dot join. If the trend clearly changes (e.g. peak then decline), draw two separate best-fit lines. Do NOT join the last point back to the origin unless it actually belongs there
- Anomalous points: circle and label anomalous points; do NOT include them when drawing the line of best fit
Key calculations
Rate of reaction = gradient of a tangent to the curve at the point of interest.
For a straight-line portion: gradient = Δy / Δx (choose two points far apart on the line; use the coordinates to calculate).
Units of rate: if y = volume (cm³) and x = time (min), rate = cm³ min⁻¹.
Mean: ∑x / n
Range: maximum − minimum value
Standard deviation (SD): measures spread of data around the mean. Large SD = wide spread; small SD = values close to mean. Calculated using calculator or formula (Paper 5 provides formula if needed).
SD is preferred over range because it uses all data points, not just extremes.
% change = [(final − initial) / initial] × 100
A negative value indicates a decrease. Always specify the direction ("increased by 12%", "decreased by 8%").
Serial dilution: to make 1% from 10% stock: mix 1 part stock + 9 parts water. Each step reduces concentration by the dilution factor.
To achieve specific concentrations: C₁V₁ = C₂V₂ (the dilution equation).
Statistical tests — the 3-tier framework
Paper 5 requires you to select and apply the appropriate statistical test. The choice depends on the type of data and the question asked:
| Test | When to use | What it tests |
|---|---|---|
| Chi-squared (χ²) | Categorical data; comparing observed vs expected frequencies (counts, not means) | Whether observed frequencies differ significantly from expected (e.g. Mendelian ratios, numbers of organisms in habitat types) |
| Student’s t-test | Normally distributed continuous data; comparing the means of two groups; sample size ≥ 5 per group | Whether the difference between two sample means is statistically significant (e.g. leaf area in sun vs shade plants) |
| Spearman’s rank correlation coefficient (rs) | Ranked / ordinal data or non-normally distributed continuous data; assessing the relationship between two variables | Whether there is a significant correlation (positive, negative, or none) between two variables (e.g. enzyme activity vs pH) |
- State the null hypothesis (H₀): "There is no significant difference / correlation between [IV] and [DV]."
- Calculate the test statistic (χ², t, or rs) using the appropriate formula; show full working
- Determine degrees of freedom (df): for t-test, df = (n₁ − 1) + (n₂ − 1); for chi-squared, df = number of categories − 1
- Compare to critical value at p = 0.05 from the statistical table provided
- Draw a conclusion: state whether to accept or reject H₀ and what this means biologically
- Rank each variable separately (1 = smallest; ties get the mean of their ranks)
- For each pair of data points, calculate d = rank of x − rank of y
- Calculate d² for each pair; sum to get ∑d²
- Apply formula: rs = 1 − [6∑d² / n(n²−1)]
- rs ranges from −1 (perfect negative) to +1 (perfect positive); 0 = no correlation
- Compare |rs| to critical value at p = 0.05 for the appropriate n
Drawing conclusions from data
- Describe the trend precisely: use data values. "As [IV] increases from X to Y, [DV] increases from A to B, suggesting a positive relationship." Not just "increases".
- Reference the data: quote specific values from the table or graph to support the trend
- Relate to the biology: explain the mechanism: "This is consistent with enzyme kinetics: as substrate concentration increases, more active sites are occupied, increasing the rate of product formation."
- Note exceptions: if a point doesn't fit the trend, acknowledge it. "The value at 50°C is anomalously low, possibly because..."
- Do NOT extrapolate beyond the data: only draw conclusions for the range of values tested
A student investigates the effect of light intensity on the rate of photosynthesis. She measures the volume of O₂ produced per minute at six light intensities. The data are shown below.
Light intensity (klux): 2, 4, 6, 8, 10, 12
Mean O₂ produced (cm³ min⁻¹): 1.2, 2.4, 3.6, 4.5, 4.8, 4.9
(a) Describe the trend shown. [3]
(b) A student claims: "The rate increases proportionally with light intensity throughout." Evaluate this claim using the data. [2]
(c) Suggest the most appropriate statistical test to assess whether the correlation between light intensity and rate is significant, and justify your choice. [1]
(a) Trend [3 marks]
- Between 2 and 6 klux, rate of O₂ production increases proportionally (linearly) from 1.2 to 3.6 cm³ min⁻¹ — a threefold increase as light intensity triples [1]
- From 8 klux onwards, the rate plateaus; only increasing from 4.5 to 4.9 cm³ min⁻¹ between 8 and 12 klux — a much smaller increase [1]
- This suggests that light intensity is no longer limiting above ~8 klux; another factor (CO₂ concentration or Rubisco activity) becomes limiting [1]
(b) Evaluate claim [2 marks]
- The claim is partially correct: from 2 to 6 klux, rate doubles as light intensity doubles (2→4 klux: 1.2→2.4; 4→6 klux: 2.4→3.6) — consistent with proportional increase [1]
- The claim is incorrect for the higher range: from 8 to 12 klux, the rate increases by only 0.4 cm³ min⁻¹ despite a 50% increase in light intensity — the rate has levelled off and is no longer proportional [1]
(c) Statistical test [1 mark]
- Spearman’s rank correlation; because both variables are continuous measurements and the question asks whether a correlation (not a difference between groups) is significant; the data do not appear normally distributed (plateau effect) so a non-parametric test is appropriate [1]
Evaluation & improvements
Types of experimental error
| Error type | Definition | Effect on results | How to reduce |
|---|---|---|---|
| Random error | Unpredictable variation; affects individual measurements differently; causes scatter | Spread of data around the mean; reduces precision | Repeat measurements; take large sample size; calculate mean |
| Systematic error | Consistent bias in one direction; affects all measurements equally; often due to instrument calibration or method flaw | All results shifted in one direction; does not affect the trend but affects absolute values | Calibrate instruments; use appropriate controls; re-examine method for bias |
| Human error | Mistakes in reading scales, timing, transferring volumes | Random or systematic depending on nature of mistake | Careful technique; digital instruments; automated timing; training |
| Zero error | Instrument does not read zero when it should | Systematic offset in all readings | Re-zero before use; correct by subtracting offset from all readings |
Identifying limitations
A limitation is a feature of the experimental design or execution that reduces the reliability, validity, or accuracy of the results. For each limitation, explain how it affects the results:
Weak (0 marks): "There was human error in the experiment."
Strong (2 marks): "Reaction time when starting and stopping the stopwatch introduces random error in the measured time; this could cause the calculated rate to be inaccurate, particularly for fast reactions where the reaction time is a larger proportion of the total time measured."
Common high-mark limitations for biology practicals:
- Biological variation between organisms (e.g. different potato discs have different catalase concentrations) — increases random error; addressed by using tissue from the same organism
- Temperature fluctuation during a reaction — systematic or random; use a thermostatically controlled water bath
- Difficulty judging a colour change endpoint — introduces subjective random error; use a colorimeter for objective measurement
- Loss of gas during measurement — systematic underestimate of O₂ produced; check all joints and connections for leaks
- Meniscus reading on a graduated cylinder — parallax error; use a syringe instead
- Enzymes begin to deteriorate over the duration of the experiment — reduces validity if comparing different temperatures; prepare fresh enzyme solution for each trial
Suggesting improvements
Structure every improvement answer as: "[Limitation]. To improve this, [specific modification]. This would [effect on results]."
- Replace stopwatch with data-logger / oxygen electrode: automated, continuous, objective recording eliminates human reaction time error and allows calculation of initial rate more precisely
- Use a buffer solution to control pH: prevents drift in pH as the reaction progresses; ensures any changes in rate are due to the IV, not pH change
- Increase the number of replicates from 3 to 5+: provides better estimate of the mean; outliers have less influence; allows calculation of a more reliable standard deviation
- Use a thermostatically controlled water bath (±0.1°C): maintains more precise temperature control than a manually topped-up bath; reduces systematic temperature error
- Use a colorimeter with a standard curve: converts absorbance to concentration objectively, avoiding subjective colour comparison
- Use tissue from the same organ of the same individual: controls for biological variation in enzyme concentration between different source organisms
Anomalous results
An anomalous result is a value that does not fit the overall trend. Paper 5 may ask you to identify anomalous values and explain how to handle them:
- Circle and label the anomalous point on the graph; do NOT include it in the line of best fit
- Do NOT include it when calculating the mean for that data point
- Suggest a possible cause: instrument malfunction, contamination, timing error, recording error
- State that the measurement should be repeated to determine whether the anomaly persists or was due to random error
Extending the investigation
Paper 5 questions often ask how the experiment could be extended. Effective extensions:
- Widen the range of the IV to determine the full dose-response curve (e.g. test temperatures from 0°C to 80°C not just 20–60°C)
- Narrow the range around the interesting region (e.g. around the optimum temperature) to identify the precise optimum more accurately
- Test additional variables (e.g. having established the effect of temperature, also investigate the effect of pH)
- Test different organisms or enzyme sources to determine whether the relationship is specific or general
A student investigated the effect of temperature on the rate of breakdown of starch by amylase. She used iodine solution to test for the presence of starch every 30 seconds. The time taken for the starch to fully break down was recorded. She found that at 20°C, starch took 4.5 minutes; at 30°C, 2.0 minutes; at 40°C, 1.2 minutes; at 50°C, 3.8 minutes; at 60°C, starch was never fully broken down.
(a) Identify ONE limitation of using iodine solution to measure starch breakdown, and suggest how this could be improved. [2]
(b) The student used a different bottle of amylase solution for the 50°C trial. Explain why this is a source of error and how it could be controlled. [2]
(c) The result at 50°C seems unexpectedly high. Suggest ONE possible explanation. [1]
(d) State how the student could determine the precise optimum temperature more accurately. [1]
(a) Limitation of iodine method [2 marks]
- The endpoint (colour change from blue-black to colourless) is subjective — different observers may judge the endpoint differently, introducing random error in the recorded time [1]
- Improvement: use a colorimeter set to ~660 nm to measure absorbance at fixed time intervals; a calibration curve can convert absorbance to starch concentration, giving a quantitative, objective measurement [1]
(b) Different amylase bottle [2 marks]
- Different bottles of amylase may have different enzyme concentrations (or the bottle may have partially denatured during storage), meaning the rate at 50°C cannot be directly compared to the rates at other temperatures [1]
- Control by using a single stock solution of amylase for all temperatures; prepare all aliquots from the same bottle at the start of the experiment [1]
(c) Explanation for anomalous result at 50°C [1 mark]
- The different amylase bottle may have had a lower enzyme concentration; or partial denaturation of amylase above the optimum temperature (~35–40°C) has begun, reducing the rate; or the pH of the 50°C solution may have shifted [1]
(d) More accurate optimum temperature [1 mark]
- Narrow the temperature range around the apparent optimum (~40°C); test at smaller intervals (e.g. 38, 39, 40, 41, 42°C) to pinpoint the temperature at which breakdown is fastest [1]
Paper 5 Practice — Integrated Questions
Full Paper 5-style integrated questions covering planning, analysis, and evaluation.
A researcher compares the mean leaf area of the same plant species grown in full sunlight versus deep shade. Each sample contains 30 leaves. Which statistical test should be used to determine whether the difference in means is significant?
- A. Chi-squared test — compares observed and expected frequencies
- B. Student’s t-test — compares the means of two groups of continuous data
- C. Spearman’s rank correlation — tests for a correlation between two ranked variables
- D. Simpson’s diversity index — measures species diversity
A student investigates the effect of NaCl concentration on water uptake by potato cylinders (osmosis).
(a) State the hypothesis for this investigation. [2]
(b) The student measures the percentage change in mass of potato cylinders after 30 minutes in different NaCl concentrations (0, 0.1, 0.2, 0.3, 0.4 mol dm⁻³). State the DV and describe how it is measured. [2]
(c) State TWO variables that must be controlled and explain how each is controlled. [4]
(d) The student calculates % change in mass at 0.3 mol dm⁻³ as +2.5%, but on repeating obtains −0.5%. Identify the type of error this represents and suggest how to address it. [2]
(a) Hypothesis [2 marks]
- As NaCl concentration increases, the percentage change in mass of potato cylinders will decrease (from positive to negative), because increasing external osmolarity reduces the water potential gradient between the potato cell sap and the external solution [1]
- At the potato’s isotonic point, there will be no net movement of water and no change in mass; above this concentration, mass will decrease as water moves out by osmosis down the water potential gradient [1]
(b) DV and measurement [2 marks]
- DV: percentage change in mass of potato cylinder [1]
- Measured by: blot each cylinder dry with tissue paper to remove surface water; weigh before (m₁) and after (m₂) immersion using a balance (to 2 decimal places); calculate % change = [(m₂ − m₁) / m₁] × 100 [1]
(c) Two controlled variables [4 marks; 2 each]
- Temperature: temperature affects the rate of osmosis by changing membrane fluidity and water molecule kinetic energy; control by conducting all trials at the same room temperature or in a water bath set to 25°C [2]
- Surface area / dimensions of potato cylinder: larger surface area increases the rate of osmosis; cut all cylinders to the same length and diameter using a cork borer and ruler (e.g. 40 mm × 10 mm); check mass is consistent before treatment [2]
(d) Error type and solution [2 marks]
- The large discrepancy between replicates (+2.5% vs −0.5%) represents random error — possibly from inconsistent blotting of surface water, different cylinder masses, or contamination [1]
- Repeat the measurement at 0.3 mol dm⁻³ at least once more; calculate the mean of the three trials; investigate the source of inconsistency (ensure consistent blotting technique and identical cylinder sizes) [1]
Paper 5 — Common Mistakes
- 📋Vague variables without units or range"Temperature" is incomplete. "Temperature, varied from 20°C to 60°C in 10°C steps" earns the mark. Always: name + units + range or method of measurement. Same applies to DV: "rate" alone is wrong; "initial rate of O₂ production in cm³ min⁻¹" is correct.
- ✏Method steps that cannot actually be followedMethods must be executable. "Measure the substance" tells a reader nothing. "Use a 1 cm³ syringe to transfer exactly 1.0 cm³ of 0.5% starch solution into a test tube" is executable. If you cannot picture yourself doing it in a lab, the step is not specific enough.
- 📉Joining data points dot-to-dot instead of drawing a line of best fitConnect-the-dots lines follow every noise point and misrepresent the underlying trend. A best-fit line (smooth curve or straight line) shows the underlying relationship. Draw through the majority of points with roughly equal numbers above and below. Label anomalous points and exclude them.
- 🔎Confusing random and systematic errorRandom error affects individual measurements unpredictably (varies between replicates). Systematic error shifts all results in the same direction. Random error is reduced by repeating and averaging; systematic error is not reduced by repeating — it requires identifying and correcting the source of bias.
- 📊Selecting the wrong statistical testChi-squared = categorical frequency data (counts). t-test = comparing two means of normally distributed continuous data. Spearman's = testing correlation between two variables (especially if data are not normally distributed). Getting the choice wrong typically loses 2–3 marks in the analysis section.
- 📓Null hypothesis is the wrong way round or too vagueThe null hypothesis must state "no significant difference / no significant correlation." Never state "there IS a difference" as H₀ — that is the alternative hypothesis. And specify the variables: "There is no significant difference between the mean leaf areas of sun-grown and shade-grown plants" not just "there is no significant difference."
- 👤Stating "human error" as a limitation without being specific"Human error" earns no marks by itself. You must identify the specific action prone to error: "parallax error when reading the meniscus of a measuring cylinder" or "reaction time when starting the stopwatch at the moment of mixing" — and explain its effect on the results.
- 🖥Axes labelled without units, or IV and DV reversedIV always on the x-axis; DV on the y-axis. Both axes must be labelled with "Variable name / units" (e.g. "Concentration / mol dm⁻³"). Missing units on axes typically costs 1 mark per axis in graph questions — check every axis before moving on.
- 🔍Conclusion says "proves" or "proves beyond doubt"Science does not prove. Statistical tests show that results are "consistent with" or "significantly different from" the null hypothesis at a given probability level. Use: "the data suggest," "the results support the hypothesis," "there is a significant [positive / negative] correlation." Never "proves."
Important: Formal statistical tests (t-test, Spearman, chi-squared, Simpson's index) may or may not be required in any given Paper 5 question. Paper 5 heavily tests experimental design, graph interpretation, limitations, validity, reliability, and improvements — these skills appear in every paper. Statistical tests are one tool among many; always read the question carefully before assuming a test is needed. When a test is required, the question will ask explicitly (e.g. “name an appropriate statistical test” or “justify your choice of statistical test”).
Planning: hypothesis with IV + DV + biological reason; IV with range and units; DV with measurement method and units; ≥2 controlled variables with how; control experiment; ≥3 replicates; safety point.
Analysis: table with headings including units; mean column; graph with IV on x-axis, DV on y-axis, both labelled with units, appropriate scale using >50% of grid; best-fit line not dot-to-dot; conclusions with data values and biological explanation.
Evaluation: limitations are specific and linked to an effect on results; improvements are specific and actionable; anomalous results identified, circled, and excluded from line of best fit; statistical test selected correctly with justification; null hypothesis stated precisely.