Choosing and justifying sampling techniques for an investigation

Overview

What justifying a sampling technique means rather than describing it, worked through a beach, a questionnaire, a dune succession and an urban transect, and how sample size is settled.

What you need to know

Answer each one before you reveal it.

Questions from this subtopic

Each one with the model answer underneath.

1. 1

   What does it mean to justify a sampling technique, rather than just to describe it?

   Model answer

   Describing says what you did - a quadrat every ten metres along a transect. Justifying says why that was the right decision for this investigation and why the alternatives were rejected: the hypothesis concerns change with distance, so even spacing along the gradient captures it; random placing risked clustering and leaving gaps in the succession; stratification was impossible because the zones were not known in advance. A justification also acknowledges what the chosen method costs you, since every method has a weakness that has to be lived with.

2. 2

   How would you justify the sampling for an investigation into pebble size along a beach?

   Model answer

   Use a quadrat, because pebble size must be sampled over a known area rather than by picking pebbles up, which invites unconscious selection of the biggest and most attractive. Place the quadrats systematically at fixed intervals along a transect if the hypothesis concerns change with distance along or up the beach, so the gradient is evenly covered; place them randomly within each zone if the aim is to compare zones fairly. Measure every pebble within the frame, with an agreed rule for those on the edge, and state the frame size so the sites are comparable.

3. 3

   How would you justify the sampling for a questionnaire measuring a town's sphere of influence?

   Model answer

   Sample at several sites around the centre rather than one, so that customers of different shops and users of different car parks and bus stops are reached, and sample at more than one time of day and on more than one day of the week, since weekday and weekend shoppers travel from different distances. Within that, approach respondents systematically, for example every fifth person passing, so the surveyor cannot choose who looks approachable. If the age structure of the users is known and matters, stratify the target numbers by age band. State the sample size and admit that non-response biases any street survey.

4. 4

   How would you justify the sampling for a vegetation survey across a sand dune succession?

   Model answer

   Use quadrats, because vegetation is measured as cover or frequency per unit area. Place them along a line transect running from the strand line inland, because the hypothesis concerns change with distance from the sea and the transect follows that gradient. Space them systematically so no part of the succession is missed, or stratify them by dune zone if the aim is to compare embryo, yellow and fixed dunes with equal numbers in each. Repeat with several parallel transects so that a single atypical line, such as one crossing a blowout or a path, does not determine the conclusion.

5. 5

   How would you justify the sampling for a land use survey along a transect out of a city centre?

   Model answer

   A line transect suits the hypothesis, because the prediction concerns change with distance from the centre, and the route should follow a main road outwards so that it crosses the successive zones. Record land use systematically at fixed distance intervals, or record every property within fixed-length sections, so that the sampling effort is the same at every distance. Use more than one transect, ideally in different directions, because a single route may follow an unusual corridor such as an industrial estate or a park and would then misrepresent the city as a whole.

6. 6

   When is pragmatic sampling defensible, and how should it be reported?

   Model answer

   It is defensible when the alternative is no data at all or an unacceptable risk: when random points fall in deep or fast water, on a cliff face, on private land where permission has been refused, or where the tide or traffic makes access impossible. It is reported honestly - stating that sites were selected on grounds of safe access rather than by chance, saying how many sites were relocated and why, and discussing in the evaluation how that is likely to have biased the results, for example towards shallower, calmer or more accessible parts of the study area.

Other ways to work this

The same content, four other ways.

What is inside Levo

Six ways to work the same material.