5(c)–(e); 3i Section 2 - Field methods, sampling and ethical collection

Matching field methods to questions

A method is appropriate when the information it produces can help answer the question. Measuring a beach profile can establish slope and changes in cross-sectional form; pebble measurements can investigate sediment characteristics; interviews can explore how residents experience management. They are not interchangeable indicators of a single concept. A defensible design explains the geographical role of each method rather than collecting as many different data types as possible.

Define measurements and categories before collection. A beach-profile procedure needs consistent slope intervals, reference points and sign conventions; sediment measurements need an agreed axis and selection rule. A pedestrian count needs a clear boundary, duration and treatment of repeated crossings. An environmental-quality scale needs defined criteria so “poor” or “good” is not interpreted differently by each observer. If the concept is subjective, retain that subjectivity as part of the evidence rather than disguising a judgement as an instrument reading.

A pilot tests whether the procedure produces interpretable, relevant data. It may reveal that a question is leading, a category is ambiguous or an instrument's resolution is insufficient. Revise the method for a stated reason and record the change. Pilot data should not be silently merged with later data if the method changed their meaning.

Use complementary methods when they answer distinct needs. An illustrative access enquiry might combine mapped routes, timed journeys and interviews about barriers. These can reveal disagreement: a short route may still be inaccessible to a wheelchair user. Such disagreement can improve the geographical explanation. It does not necessarily mean one method failed.

Practical knowledge includes using equipment correctly, recognising limitations and following centre-approved safety and environmental procedures. The investigation needs a record of fieldwork actually performed, including how methods were used and adapted.

Hayhoe and colleagues' 2011 study of Upper Xingu headwaters used stream measurements and rating-curve calibration during 2007–2009 to compare forest and soya catchments. The study shows the connection between question, instrument and variable: water-level observations need a calibrated relationship to become discharge estimates. Choosing that method has a hydrological purpose, while calibration and catchment comparability limit interpretation.

Sampling, frequency and timing

Sampling and timing determine which part of a geographical system becomes visible. Explain the population, sampling frame, selection rule and sample size in relation to the question. Random, systematic and stratified designs have different strengths and limitations; naming one is not a justification. For example, stratifying a coastline by lithology may help compare rock influences, while a systematic transect may reveal a continuous gradient.

Spatial sampling alone may miss temporal variation. A drainage-basin measurement can vary with recent rainfall; beach form with tides and storms; urban use with weekday, hour and season. Decide whether the purpose concerns a typical condition, a comparison at matched times or a changing process. Repetition should capture that purpose rather than add convenient duplicates. Many measurements taken seconds apart at one site do not automatically constitute independent evidence about many places.

Frequency and duration need justification. Counts collected over unequal periods require conversion to common rates; otherwise activity is confounded with observation effort. Sediment sites measured at different tides may similarly represent different beach zones.

Worked boarding rates

At one bus stop 84 people board in seven minutes; another records 126 in 14 minutes. Divide each count by its own duration:

The larger total at B represents a lower rate. Comparable periods and service schedules still matter.

Worked cycle-crossing rates

One crossing records 156 cycles in 12 minutes, and another 96 in eight minutes:

Rates remove the duration difference, but repeating comparable weekday and weather conditions is needed before generalising about typical use.

Practical constraints matter, but describe their effects. A closed footpath may require a revised frame; a missed site should not be replaced with an apparently similar one without recording the decision. Explain how the change affects coverage and interpretation. Pilot results can inform a feasible sampling density and identify whether the proposed method is sensitive enough to detect differences.

Justification is therefore a reasoned link between method and enquiry, not a claim that the design is perfect. Identify the relevant source of variation, explain how the selected design captures it and acknowledge what remains outside its scope. This makes the eventual evaluation specific and evidence-based.