Small Group Tutorials

Here to help students catch up, keep up, and move ahead. Book a consultation here.

How to Perform in the new G3 SEC Examinations | Learner’s Guide Vol 1475 | Biology Orchard Leaf Survey — Stomata, Microscope Counts and Sampling Bias

Quick information: This 2027 SEC G3 Biology K325 learner guide develops the ability to interpret a fictional school-orchard leaf survey. The focus is not a general account of photosynthesis or plant transport, already covered in earlier guides. It is a closer examination of leaf structure, stomata, equal-area microscope fields, sampling, descriptive statistics and the difference between an observation and a biological explanation.

Direct answer for parents: in this original data set, four shaded-leaf microscope fields contain 8, 10, 9 and 11 counted stomata, while four sun-exposed-leaf fields contain 13, 12, 15 and 14. Their means are 9.5 and 13.5 stomata per equal-area field. The counts suggest a difference in these particular sampled fields. They do not, by themselves, prove different transpiration rates, a causal effect of sunlight, a difference between whole orchards or anything about actual Sengkang trees.

At a glance:

  • Check the 2027 G3 Biology K325 theory and practical-paper structure.
  • Distinguish stomata, guard cells, mesophyll and the functions of a leaf.
  • Calculate means and densities from fictional equal-area microscopic observations.
  • Detect pseudoreplication, position effects, observer bias and unsupported causal claims.
  • Practise structured explanations, data handling, fair tests and honest conclusions.
  • Continue from orchard Biology Vol 1463 and the EMS evidence review Vol 1474.

1. The question behind the leaf survey

A student looks through a microscope and sees small pores on a leaf surface. Counting pores seems simple until the class compares two leaves. One sample comes from a sunnier part of the fictional orchard; another comes from shade. The number of pores in the viewed fields differs. What does that difference mean? This is where a student must slow down. A number is an observation. A cause is an explanation that requires enough evidence.

This chapter trains the passage from microscopic observation to carefully qualified interpretation. The reader will practise distinguishing a stomatal pore from the guard cells that regulate its aperture, identifying the reason leaf structure supports gas exchange, recording counts consistently, calculating means and comparing what the records can and cannot show.

The activity is not a prescription to collect leaves from public property. Any actual microscope exercise needs school permission, appropriate supervision and a method approved by teachers. All figures here are invented learning data. There is no claim that these counts were collected at a real orchard.

2. The current 2027 official Biology specification

SEAB identifies G3 Biology for the 2027 SEC as K325, with the older reference code 6093 shown in its syllabus directory. The official K325 scheme requires Papers 1, 2 and 3. Paper 1 is 40 compulsory multiple-choice items in one hour, carrying 40 marks and 30% of the assessment. Paper 2 is structured and free-response work over one hour 45 minutes, carrying 80 marks and 50%. Paper 3 is the practical examination over one hour 50 minutes, carrying 40 marks and 20%.

The practical assessment includes planning, manipulation and measurement, presentation of data, analysis, conclusion and evaluation. These are exactly the habits the survey develops, although the imaginary survey is not an authentic practical examination and does not imply that every candidate will be asked to count stomata. The syllabus also covers plant structure, photosynthesis, transpiration and translocation. Students should keep subject content and practical reasoning connected.

A learner taking combined Science should check the relevant code and assessment pattern instead of assuming pure Biology’s K325 details apply to them. The title ‘SEC’ refers to the qualification from 2027; the level and chosen subject determine the syllabus.

3. Locate the stomata in a useful mental picture

A typical dicotyledonous leaf has an upper epidermis, mesophyll tissues, vascular tissue and a lower epidermis. The palisade mesophyll contains many chloroplasts that capture light for photosynthesis. The spongy mesophyll contains air spaces that facilitate gas movement between cells and the internal atmosphere. Stomata are pores commonly found in the epidermis through which gases and water vapour can move.

A pair of guard cells surrounds a stomatal pore and helps regulate opening. When the pore is open, carbon dioxide can diffuse into a leaf and oxygen can diffuse outward according to concentration gradients. Water vapour may also leave through stomata, linking gas exchange with transpiration. The exact movement at any time depends on environmental and physiological conditions; a diagram should not become a claim that gas exchange always proceeds at one fixed rate.

For an exam diagram, label structures precisely. Do not call the entire pair of guard cells a pore. Do not describe xylem vessels as the location of photosynthesis. Explain how leaf structures support functions rather than memorising an unlabeled drawing.

4. Keep pore number separate from pore opening

A field with more stomata contains more counted pores under the specified observation method. That count alone does not tell us how wide the pores are open, whether they are open at the time of observation or the actual rate of water vapour loss. A student who sees 14 stomata in one field and eight in another cannot conclude that the first leaf transpires exactly 75% faster.

Stomatal conductance and whole-leaf transpiration depend on several influences, including opening, environmental conditions and leaf properties. The current data pack records only counts in equal areas. It does not record opening width, air movement, temperature, relative humidity or water loss. Those missing quantities matter when moving from structural data to a rate claim.

The disciplined conclusion is: ‘The selected fields differ in counted stomatal number.’ A proposed biological explanation may be discussed as a hypothesis, but it should not be presented as a demonstrated cause. This distinction is as important as the arithmetic.

5. The original data pack

The fictional teaching sheet uses microscope fields of equal known area, each 0.20 square millimetres. Four non-overlapping fields from one shaded leaf have stomatal counts 8, 10, 9 and 11. Four non-overlapping fields from one sun-exposed leaf have counts 13, 12, 15 and 14. Assume the same magnification, field area, counting rule and competent observer were used for both sets.

The sheet intentionally includes only one leaf from each condition. That limitation will become important. Four fields are four measurement positions, but not four independently sampled trees or four separate leaves. A student must not mislabel the sample size to make a stronger claim. The words ‘one leaf per condition’ should be copied into the methods summary.

The data are deliberately tidy enough for arithmetic but not strong enough for a sweeping causal conclusion. Students often earn insight by understanding why a small example can be mathematically clear yet biologically limited.

6. Compute the means, one operation at a time

For the shaded leaf, add 8 + 10 + 9 + 11 = 38. Divide by the four viewed fields to obtain a mean of 9.5 stomata per field. For the sun-exposed leaf, add 13 + 12 + 15 + 14 = 54. Divide by four to obtain 13.5 stomata per field. The difference between the sample means is four stomata per field.

That difference is valid for these records. It does not express a species-wide, tree-wide or orchard-wide difference. It is also not correct to say that the sun leaf has ‘four more stomata in total’. The observed total counts were 54 and 38 across the selected microscope fields, while four is the difference between the means per equal-area field.

To check the arithmetic, multiply 9.5 by four and recover 38, then multiply 13.5 by four and recover 54. This reverse calculation catches common transcription mistakes before the explanation is written.

7. Convert equal-area counts into density

Each field covers 0.20 mm². The shaded-leaf mean density from the selected fields is 9.5 ÷ 0.20 = 47.5 counted stomata per square millimetre. The sun-leaf mean density is 13.5 ÷ 0.20 = 67.5 per square millimetre. Because the same area was used for both leaves, the ratio of their estimated densities equals the ratio of their mean counts.

State the units. ‘67.5 stomata’ is not the same quantity as ‘67.5 stomata per mm²’. A density allows comparison across areas, but the numerical estimate still depends on sampling and counting accuracy. It should be described as an estimate from observed fields, not the exact density of every part of the leaf.

If the microscope fields differed in area, raw counts could not be compared fairly without normalisation. A smaller field might contain fewer stomata even if its underlying density were higher. Identify the denominator before interpreting a raw count.

8. Work out the percentage comparison responsibly

The difference between the two mean counts is 13.5 – 9.5 = 4. Relative to the shaded-leaf mean, that is 4 ÷ 9.5 × 100 ≈ 42.1%. The sentence should therefore say ‘The sun-exposed leaf’s sampled mean is about 42% higher than the shaded leaf’s sampled mean, using the shaded-leaf value as the reference.’

Changing the denominator changes the percentage. Relative to 13.5, the same difference is about 29.6%. Neither figure should be written without naming the baseline. Students frequently lose the meaning of a percentage by stating it as though the base were obvious when it is not.

A stronger biological answer may not need any percentage at all. It could report the two means and explain the sampling limitation. Use additional calculations when the question requests them or when they make an inference clearer, not merely to display numerical sophistication.

9. What can reasonably be concluded

There is a difference in the counts observed in the selected equal-area fields. The observed sample means are 9.5 and 13.5 stomata per field. That is the secure conclusion. It is not secure to say that bright light caused a 42% increase in stomatal density, because the source does not establish that exposure was the only difference between the sampled leaves or that the leaves are representative.

An examiner-style response can say: ‘The sun-exposed sample has a higher mean count per equal-area field, but each condition is represented by one leaf and the study has not controlled leaf age, plant identity and other influences. Further independently sampled leaves would be needed before attributing the difference to light conditions.’ The conclusion uses the evidence while recognising limitations.

Be cautious with the term ‘significant’. In ordinary conversation it can mean large or important; in statistics it often implies a tested inference. No inferential test has been supplied here. It is safer to write ‘higher in this sample’ rather than ‘significantly higher’ unless a valid statistical analysis justifies the latter.

10. Pseudoreplication: four fields are not four plants

Suppose a pupil writes, ‘We tested four shade plants and four sun plants.’ That is false. The source says one leaf from each condition, with four microscope fields per leaf. A microscope field is a sampling location nested within a leaf. Fields can reveal variation across the leaf, but they are not independent biological replicates of the light environment.

This is a sophisticated but teachable idea. Four photographs of one apple do not amount to four apples. Four measurements of one student do not amount to four students. Repeated readings help assess measurement consistency; independent biological samples help assess how generally the relationship holds. Both can be valuable, but they answer different questions.

A better next study would sample several suitable leaves from more than one plant under each defined condition, using an appropriate scheme. Students should explain why that improves representation, while recognising that sampling and experimental design still need to be fair.

11. Sampling position is not a small detail

Stomatal density may vary across different regions of a leaf, between upper and lower epidermal surfaces, with leaf maturity and among species. The original data do not say whether every field came from a comparable position. If students preferentially choose areas where pores are easiest to see, the selection may favour certain counts and bias the average.

A fair protocol might define the leaf surface and a consistent region, avoid damaged areas, use a random or systematic selection rule, maintain a known field area and record any deviations. The precise classroom technique should be supplied and supervised by the teacher. Consistency reduces some differences between samples, while truly representative sampling tackles another problem.

Do not confuse ‘same magnification’ with ‘same area’ unless the instrument and field of view are also known. A worksheet should state or establish the observed area rather than assuming it follows from a magnification number alone.

12. The counting boundary rule

When a stomatal pore lies partly at the edge of a microscope field, a class needs an agreed rule to avoid arbitrary inclusion. One common approach in microscopy counting is to include objects touching specified edges and exclude those touching the opposite edges. The exact rule is less important for this learning exercise than applying it consistently to every field and documenting it.

Without a boundary rule, different observers may include or exclude partly visible stomata in different ways. That introduces observer variation. If students disagree about a count, they should inspect whether they used the same definition and the same field boundaries before concluding that one observer is careless.

Counting reliability is not merely a matter of attention. It is influenced by definitions, field selection, image clarity and recording. A precise method helps honest observers produce comparable results.

13. Why the leaf’s two surfaces matter

The upper and lower epidermal surfaces may have different stomatal distributions. Many terrestrial dicotyledonous leaves have more stomata on the lower surface, but this is not a universal rule for all plant species and environments. An answer about a particular specimen should rely on the labelled observation or the conditions supplied in the question.

If a class mixes counts from upper surfaces in one condition with lower surfaces in another, the comparison is confounded. Even if the arithmetic is flawless, the result could reflect surface differences rather than light exposure. Good experimental design makes sure the compared samples represent the intended variable.

When drawing a cross-section, pupils should connect upper and lower epidermis, mesophyll, stomata and vascular tissues to their functions. Names in a diagram are useful only when students understand their relationships.

14. Connect stomata to photosynthesis without overstating

Carbon dioxide enters leaves mainly through stomata and diffuses to photosynthesising cells. Light is needed for photosynthesis; chloroplasts in mesophyll cells absorb energy. The rate of photosynthesis can depend on light intensity, carbon dioxide concentration and temperature, among other factors. These limiting-factor ideas are in the official 2027 Biology syllabus.

Yet the number of stomata visible in a field does not give a direct measurement of carbon dioxide uptake or photosynthetic rate. Pores may have different apertures. Other factors may limit the process. A student should not use a structural count as a substitute for a gas-exchange experiment without evidence.

An appropriate bridge sentence is: ‘Stomata provide pathways for gas exchange relevant to photosynthesis, but the observed stomatal count alone does not measure the rate of photosynthesis.’ This acknowledges biological function while respecting the data’s scope.

15. Connect stomata to transpiration carefully

Transpiration is the loss of water vapour from aerial parts of a plant, especially through stomata in leaves. Its rate can be affected by air movement, temperature, humidity and light intensity. The syllabus expects learners to investigate and explain relevant factors, and to understand how transpiration contributes to water movement through plants.

A student who remembers these facts may be tempted to say the sun-exposed leaf with more counted stomata must lose more water. That is not established here. The leaf areas, pore openings, environmental conditions and actual water loss are unmeasured. Density and rate are distinct. Even if a structural feature influences a process, the feature alone need not determine its observed speed.

The secure explanation is biological rather than speculative: stomata are important pathways for gas exchange and vapour loss; actual transpiration depends on more than their number.

16. Plan an improved comparison

A class could first specify a research question: ‘Do independently sampled leaves from two clearly defined environments differ in estimated stomatal density?’ It should then decide how to control or record species, leaf age, sampling surface, microscope area and field-selection rules. Several biological replicates per condition would reduce dependence on one unusual leaf.

The method should include a consistent counting definition, independently chosen microscope fields within each leaf, a transparent data table and an appropriate summary by leaf as well as by condition. The student would then interpret the results without presuming the hypothesis is true. If the data do not support a clear difference, reporting that honestly is not a failed investigation.

The improvement is conceptual: measure the quantity that the question names, replicate at the correct biological level and control competing explanations. Teachers should determine practical procedures and safety rather than pupils improvising field methods from the internet.

17. The difference between a controlled test and an observational study

Our two-leaf fictional comparison is observational. It records counts from leaves described as shaded and sun-exposed; it does not show that the investigator randomly assigned otherwise comparable plants to controlled light conditions. Therefore a causal conclusion about light is not supported. The environment might differ in humidity, plant age or other factors.

A controlled experiment would require an appropriate design, carefully defined independent and dependent variables, control of relevant conditions and sufficient replicates. Even then, conclusions should match the scope of the design. A student need not memorise sophisticated statistical vocabulary to understand this: if several things change at once, one cannot confidently assign the observed outcome to only one of them.

Examination performance improves when the student can say ‘This comparison shows an association in the sampled material’ and then identify what extra design would allow stronger causal inference.

18. How to draw a fair data table

Write a clear title describing counted stomata per 0.20 mm² field. Include columns for sample identity, microscope field number and count. Shade fields 1–4 have 8, 10, 9, 11; sun fields 1–4 have 13, 12, 15, 14. Keep the individual measurements visible and place means or density estimates in a separate summary. Avoid hiding variation by showing only a final average.

If the table uses a combined heading for the area, do not forget its unit. If a graph is drawn, label axes and use a sensible scale. Individual points or dot plots can show variation better than a pair of unexplained bars. A bar chart may compare the two means, but it should not make the viewer think a large biological sample exists.

Students should know the difference between a graph of recorded data and a graph of a modelled prediction. Neither becomes a causal proof simply through neat presentation.

19. An original multiple-choice challenge

Question: A student finds 13.5 and 9.5 mean stomata per equal-area microscope field on two leaves. Which conclusion is best supported? A: Every sun leaf has exactly four more stomata. B: In the sampled fields, the sun-leaf mean is four higher. C: Sunlight definitely caused an increase in density. D: The sun leaf loses 42% more water each minute.

The correct answer is B. A generalises from selected fields to every leaf. C asserts a cause without a controlled comparison. D substitutes stomatal counts for measured transpiration rate. B names both the quantity and the sample from which it came.

The purpose of this question is not to teach pupils to guess the cautious-sounding option every time. It is to identify which claim follows from the supplied observations and which claims require additional measurements or experimental controls.

20. A structured-response challenge

A structured question asks: ‘Calculate the mean and estimated stomatal density for each leaf, and suggest two reasons why the comparison does not establish a causal effect of light.’ A correct calculation gives 9.5 and 13.5 stomata per field, equivalent to 47.5 and 67.5 stomata/mm² when the field is 0.20 mm².

A good explanation then identifies the biological sampling problem and another uncontrolled variable. For example, each condition uses only one leaf, and leaf age or surface might differ. The student should avoid vague responses such as ‘the test was unfair’ without naming a possible source of bias. A reason is stronger when it explains how the problem could affect the inference.

This is a practical example of combining numerical data handling with biological evaluation. Show working clearly, label units and connect criticism to the actual method.

21. A practical-planning challenge

A teacher asks pupils to plan a study that estimates the stomatal density of leaves collected under authorised school procedures. The answer should state a clear question, define the specimen and surface, identify the known field area, explain field selection, set a consistent boundary rule, record the counts and calculate density. It should also include repeated fields and independent biological samples where comparison is intended.

Evaluation should name plausible limitations. For instance, one leaf might not represent a population; field positions might be selectively chosen; ambiguous boundaries could produce counting discrepancies. Improvements should match those limitations rather than giving generic instructions like ‘be more accurate’. An appropriate improvement to biased field selection is a documented random or systematic sampling method.

The response should not recommend unsupervised chemical preparations or cutting tools. School practical work is carried out with a teacher-approved method and equipment.

22. Mistakes students make with microscopy

The first mistake is using a count as though it were a density. The second is leaving out the field area and therefore missing the denominator. The third is mixing units such as mm² and cm² without conversion. The fourth is counting a guard-cell pair as two separate stomata, changing the counting definition. The fifth is forgetting the field-edge rule and making counts inconsistent.

Two more errors concern interpretation. Students may treat repeated fields on one leaf as independent leaves. They may also translate a higher stomatal count directly into a faster photosynthesis or transpiration rate. Each mistake has a specific repair: clarify definitions, check units, label biological replication and state what was actually measured.

Keep a correction table with the error, its cause and a better method. Students learn more from diagnosing one recurrent error than from copying ten perfect examples.

23. A revision routine for an uncertain learner

Start with vocabulary and identification: pore, guard cell, epidermis, mesophyll, gas exchange, transpiration and density. Then practise one arithmetic chain using four counts and a field area. Once the learner can explain those without prompts, introduce the sampling problem. Do not demand a sophisticated evaluation while the student still confuses density with rate.

Use a five-step routine: name the measured quantity, inspect the sample description, calculate with units, state a conclusion limited to the sample, and name a reason not to generalise. After several correct repetitions, change the data or the sampling description. If the child still repeats a causal claim, ask, ‘What did we actually measure?’ rather than only saying the answer is wrong.

Progress should be visible in explanations, not only marks. A child who can recognise an unsupported conclusion has become a more independent scientific reader.

24. Parents’ guide to asking useful questions

Ask the child to explain the difference between four microscope fields and four separate leaves. Ask how the field area changes a density calculation. Ask whether observing more pores proves a faster loss of water vapour. Ask what must be controlled for a fair comparison and how a student could choose fields without favouring high counts.

These questions can be answered in plain language. A parent need not know every detail of plant anatomy to detect whether the child has separated observation from inference. If the learner struggles, have them draw a rough text-labelled diagram and then explain one function at a time.

Avoid turning the exercise into a contest of memorised terminology. The goal is to make the student accurate when faced with unfamiliar observations and ordinary data.

25. Where this sits in a longer learning route

Earlier orchard Biology guides explained sugar transport through xylem and phloem, while earlier greenhouse Biology lessons covered photosynthesis and environmental factors. This guide owns a different intent: interpreting a microscopic structural survey and its sampling limits. The next step in scientific maturity is not another list of plant parts, but a better ability to say what evidence a particular measurement can support.

Students can transfer this method to counts of cells, organisms in a quadrat or particles in repeated fields, provided the subject-specific methods and assumptions are understood. The general habit is transferable; the biological details are not interchangeable.

An excellent learner should emerge able to explain the observation, calculate transparently and decline an attractive but unsupported causal story. This combination is central to science literacy.

26. Official sources and earlier owners

Use the SEAB 2027 G3 school-candidate syllabus directory and the official G3 Biology K325 syllabus for the current format and content. The data and practice questions in this article are original and fictional, not official examination tasks or findings about real plants.

Continue from Vol 1463: orchard xylem and phloem, Vol 1474: EMS water decision and Vol 1357: greenhouse photosynthesis foundations. For subject ownership, visit Sengkang Biology tuition and syllabus support, Sengkang Science practical skills and PSLE-to-secondary Science bridge.

27. Twelve diagnostic transfer questions

Use these after completing the core survey. Each commentary contains a correction or extension that should be understood, not copied as a stock phrase. Try to answer first without reading the explanation.

Diagnostic 1: Guard cells versus stomatal openings

A sketch labels two bean-shaped guard cells and the central opening. The pupil writes that the ‘guard cell is the hole’. Correct this gently: the pore is the stomatal opening, while guard cells surround and regulate its aperture. When a question asks for a structure’s function, name the correct structure and explain the contribution. A precise diagram with the wrong biological label can mislead the entire explanation.

Diagnostic 2: A microscope field changes size

Suppose a shaded field of area 0.20 mm² has ten stomata while another sample is observed in a field of area 0.10 mm² and has eight. The raw count of eight is smaller, but its estimated density is eight divided by 0.10, or 80 per mm², against ten divided by 0.20, or 50 per mm². Normalisation reverses the careless raw-count comparison. Ask which quantity the question actually wants.

Diagnostic 3: A different count on the upper surface

A student samples the lower epidermis in the sun condition and the upper epidermis in shade. Even if means differ, the sampling surface is now confounded with the light condition. It is not enough to say ‘Use the same microscope.’ The investigator must compare like with like, or deliberately study surface as a separate factor. Strong evaluation names the variable that changed unintentionally.

Diagnostic 4: When a field has no stomata

A field with zero counted stomata is still an observation if the field was selected by the agreed rule and examined properly. Do not discard it simply because it looks inconvenient. Removing low values without an established criterion can bias the mean. First check the method and recording; then retain valid data and explain the observed variation. Scientific confidence is not improved by hiding awkward results.

Diagnostic 5: Counting a boundary pore

Two pupils disagree over a pore touching the edge of the image. Before arguing about who is better at microscopy, establish the class’s inclusion/exclusion rule. Recount under the same rule and record it in the method. Repeatability depends on shared definitions. An unclear boundary rule is a correctable procedural issue rather than evidence that the biological sample changed.

Diagnostic 6: The 42% causal claim

A headline says, ‘Sunlight creates 42% more stomata.’ The calculation compared the sample means using shade as a baseline, but it did not establish light as the cause. Rewrite: ‘In this fictional survey, the sun-exposed leaf’s sampled mean was approximately 42% above the shaded leaf’s sampled mean.’ Then add the one-leaf-per-condition limitation. Wording that respects the data is more informative than a dramatic causal slogan.

Diagnostic 7: Count versus transpiration rate

A learner sees a higher density and concludes that water vapour loss must be higher by the same percentage. Ask what was measured: pore count in a fixed area, not water lost per minute or leaf area. Pore opening, environmental conditions and other factors are not included. The corrected response explains that stomata provide pathways for vapour loss, without equating density and rate.

Diagnostic 8: Four locations, one specimen

The table has four rows for each condition. A pupil confidently reports n = 4 leaves per group. Read the source description again: there is only one leaf per condition, and four fields per leaf. If the question asks for the number of independently sampled leaves, the answer is one. This distinction helps pupils understand why repeated measurements do not automatically provide biological replication.

Diagnostic 9: A mean hides position differences

Suppose four counts on a new leaf are 2, 4, 14 and 18. Their mean is 9.5, equal to the shaded-leaf mean of the original exercise. Yet their spread is much greater. Averages compress information; they do not display all variability. A prudent student reports or examines the individual results, asks whether field selection caused the spread and avoids calling two leaves identical from means alone.

Diagnostic 10: The importance of a clear question

The first draft asks, ‘Does sunlight improve the orchard?’ That is too broad for stomatal counts to answer. Rewrite the question as ‘How do the estimated stomatal densities of independently sampled comparable leaves differ between two defined light conditions?’ This is measurable. A well-defined question protects the investigation from gathering data that do not answer its actual purpose.

Diagnostic 11: An observer’s preferred fields

A student chooses only the most sharply focused, densely patterned areas while skipping blurry low-count fields. That choice can distort the result. A better protocol defines random or systematic positions and applies an objective image-quality rule before viewing the counts. Excluding unusable material may be legitimate, but the reason must be documented rather than guided by which answer looks impressive.

Diagnostic 12: Applying the evidence ladder

Write three sentences: ‘The sample mean is 13.5 per field’ is a calculated description; ‘This leaf may differ in stomatal density from the other sampled leaf’ is a limited inference; ‘Sunlight causes the difference in all orchard trees’ is unsupported. Explain the evidence needed for the third statement. The purpose is not to be afraid of conclusions, but to make each conclusion match the quality and scope of the study.