ESS IA Methods: Fieldwork, Surveys and Secondary Data

15 Sep ESS IA Methods: Fieldwork, Surveys and Secondary Data

Choose ESS IA methods by starting with the environmental issue, your research question and the evidence you can realistically collect or access. Fieldwork, laboratory experiments, questionnaires and secondary datasets can serve different questions. A method is useful when it produces relevant evidence and explains its limitations.

Choose a method that fits your question

• Fieldwork: measure environmental variables at defined locations. Explain site selection, timing, equipment, permissions and environmental safeguards.
• Laboratory work: investigate a relationship under controlled conditions. Explain which variables are controlled and how the simplified setting limits application to natural systems.
• Questionnaires: investigate reported attitudes or behaviours. Pilot the wording, protect privacy and consider who is missing from the sample. Reported behaviour is not the same as observed environmental impact.
• Secondary data: analyse an accessible dataset to answer your own focused question. Cite the source and check definitions, units, missing values, dates and whether measurements are comparable.

You do not need to claim that nobody has ever asked a similar question. Your investigation, analysis and writing must be your own. A summary of someone else’s conclusions does not replace your analysis.

Build a defensible sampling plan

There is no universal number of repeats that makes an ESS IA sufficient. Use a pilot, the variation in the data, practical constraints and the planned analysis to justify your sample. Distinguish repeated readings of the same sample from genuinely independent sampling units.

Describe where, when and how observations will be taken. Random, systematic and stratified designs have different strengths and limitations; justify the choice in context. Record calibration, instrument limitations, consent where relevant and any changes to the method.

Choose analysis by purpose, not appearance

A t-test and Spearman’s rank correlation answer different questions. A t-test concerns a comparison of means; Spearman’s correlation concerns an association between ranks. Spearman’s is not a general replacement for a t-test when data are non-normal.

• Comparing groups: consider what outcome is measured, whether observations are paired or independent and the assumptions of the chosen comparison.
• Investigating a relationship: start with a scatter plot where appropriate, inspect the pattern and consider whether a correlation method fits it.
• Categorical frequencies: a contingency-table approach may fit some questions; check expected counts and independence before using a chi-squared approximation.
• Describing a dataset: select a suitable centre and spread measure. Mean and SD are not compulsory for every variable.

Statistical significance does not establish causation or environmental importance. Explain what the analysis can support, alongside the study design and uncertainty.

Three planning examples

These are illustrations to refine with your teacher, not completed or verified high-scoring investigations. No results are supplied.

1. Vegetation near a footpath: compare defined measures of vegetation at selected distances. Consider independent transects, habitat differences and whether repeated quadrats are independent.
2. Attitudes to a plastic policy: pilot a questionnaire with defined response categories. Justify the sampling and analysis; a single Likert item should not automatically be treated as continuous data for a t-test.
3. Forest-cover change: compare a consistent dataset across defined years. Check changes in definitions and measurement methods. Two time series may move together because of a shared trend, so a simple correlation does not establish a causal link.

Explain why your method fits

Complete this planning statement in your own words: I will measure or select ___ because it can help answer ___. I will keep ___ comparable, record ___ and consider ___ as a limitation. If you cannot explain the connection between the evidence and your question, refine the design before starting.

Before collecting data

Confirm access, safety and permissions; run a pilot; prepare a data-recording sheet; and decide how the evidence will answer your question. Keep a record of raw observations and methodological decisions. Review the current ESS criteria with your teacher and use the ESS IA planning guide and IA planning examples for further context.

References

• IB ESS guide, first assessment 2026
• NIST: two-sample t-test
• NIST: Spearman rank correlation