A microscope field is usually not a separate organism
Imagine comparing stomatal counts across leaves grown under two conditions. You measure five separate microscope fields on each of three leaves in each condition. Those fields help you estimate a leaf's stomatal count more reliably. They do not necessarily provide 15 independent biological units per condition. Five fields share the same leaf, environment, genetics and handling, so they can be correlated.
First identify where the treatment was independently applied. If each of the three leaves came from an independently assigned plant, a leaf summary might be treated as one observation for the student's illustrative comparison. But if all three leaves were from a single plant in one chamber, the plant or chamber may be the true unit. The table below cannot establish independent assignment because it is simulated.
Thirty readings, six leaf summaries
All values in this section are fabricated for teaching. They are not real observations or specimen identifiers and must not be presented as collected coursework. The numbers are counts per fixed microscope field at an assumed consistent magnification.
| Condition | Leaf | Five fields | Mean per leaf |
|---|---|---|---|
| Control | C1 | 8, 9, 9, 10, 9 | 9.0 |
| Control | C2 | 11, 10, 12, 10, 12 | 11.0 |
| Control | C3 | 7, 8, 7, 8, 10 | 8.0 |
| Treatment | T1 | 15, 16, 17, 16, 16 | 16.0 |
| Treatment | T2 | 13, 14, 15, 14, 14 | 14.0 |
| Treatment | T3 | 18, 17, 19, 18, 18 | 18.0 |
Compute the mean at the level of the independent unit
The example's control leaf means are 9, 11 and 8; treatment leaf means are 16, 14 and 18. Summarize and compare these separately sampled biological units, provided your real experiment justifies that interpretation.
| Condition | Leaf means | Grand mean | Sample SD across leaves |
|---|---|---|---|
| Control | 9, 11, 8 | 9.33 | 1.53 |
| Treatment | 16, 14, 18 | 16.00 | 2.00 |
In this example the treatment mean exceeds the control mean by 6.67 counts per field. That numerical difference is not evidence that the real treatment causes an increase. Neither a small p-value calculated by falsely treating 30 nested fields as independent nor a graph without sample-unit labels can settle causation.
If leaves share plants or chambers, preserve the nesting in a hierarchical analysis or summarize at the actual independently assigned unit. State how many plants, chambers and leaves were independently assigned before choosing a significance test.
How this changes a scientific-investigation evaluation
This distinction affects three things. First, the research design should state how biological units were allocated or selected rather than just how many microscope images were collected. Second, the analysis must retain genuine independence and appropriate uncertainty. Third, the evaluation should identify the risk of shared chamber or plant effects and propose a feasible independent replication strategy.
For the first-assessment-2025 Biology scientific investigation, Research Design, Data Analysis, Conclusion and Evaluation are the four criteria, each worth six marks. No statistical method earns a mark automatically. Use the real subject guide and your specific organism, environment, permissions and controls to decide what analysis is warranted.
Five questions to ask before plotting
- Where was the treatment independently applied: chamber, plant, leaf, vessel or field?
- Which repeated values are subsamples from that same unit rather than newly randomized biological replicates?
- Are all observations equally reliable and collected with comparable field area and measurement conditions?
- Have you distinguished the number of independently assigned units from the count of raw readings in every figure?
- Would a different unit of independence materially change your claimed precision or conclusions?
The existing Biology IA pilot-study guide helps design a preliminary experiment; Biology IA examples covers whole-investigation writing. This audit isolates independence, a decision you must settle before treating subsamples as statistical replicates.
Sources & verification
Primary-source references used to verify key assessment facts on this page. Last source check: October 14, 2026.
Every count in the worked example is intentionally simulated; no claims of real experimental replication or marks are made.
- IB Biology guide, first assessment 2025Official scientific investigation guidance and Research Design, Data Analysis, Conclusion and Evaluation criteria for Biology.
- IB Biology courseOfficial current biology curriculum and assessment context.