What a field experiment is
Definition
Field experiment
An experiment carried out in a real-world setting rather than a laboratory, where the researcher still changes one variable but has less control over everything else.
- A field experiment is an experiment run in a real, everyday setting instead of a controlled one.
- You still choose and set the independent variable yourself, and you still measure the dependent variable.
- The people taking part are going about ordinary business, on a bus, in a shop queue or in a school corridor.
Example
- Suppose you want to know whether pupils hold a door open more often for someone carrying a heavy load.
- A helper walks to the same door at the same time each day, on some days carrying a tall stack of folders and on others carrying nothing, while you count who holds it open.
- The load is an independent variable you set, but who happens to walk past is entirely out of your hands.
What changes and what stays the same
- Compared with a laboratory experiment, the setting changes and the control drops sharply.
- What does not change is that a researcher still manipulates the independent variable, which is exactly what keeps it an experiment.
- If nobody set the independent variable, whatever you are looking at is not a field experiment.
Strengths
- Behaviour is more natural, because participants are in the place where the behaviour normally happens, so the findings apply far more readily to real life.
- Demand characteristics are reduced, because people often do not know a study is running and so cannot guess the aim and adjust to fit it.
Weaknesses
- Extraneous variables run loose, since weather, crowds, noise and who happens to be present can all shift the results, which also makes the study hard to repeat exactly.
- Consent is a real problem, because people who do not know they are taking part cannot agree beforehand, so you have to debrief them afterwards and offer them the right to withdraw their data.
Common Mistake
- Do not treat a natural setting as proof that a study was a field experiment.
- Do not claim a field experiment gives you the same confidence about cause as a tightly controlled one.
When to choose a field experiment
- Choose it when the behaviour would look false the moment you moved it indoors, such as helping a stranger or queueing.
- Choose it when you still want a claim about cause but need the finding to hold outside the room, and you are willing to give up control to get that.
- Choose a laboratory experiment instead when the difference you expect is small, because stray variables out in the world would swamp it.
Turning this into marks
- Give the reason as well as the label, so say that the researcher changed the independent variable in a real setting.
- Name the variable that is actually running loose in the scenario you were given.
- The mark most often lost is the bare phrase lack of control, written about field experiments in general rather than about the study in front of you.
Self review
- What still makes a field experiment an experiment?
- Explain one weakness of a field experiment and say why it matters.
- Which method would you choose to test whether wearing school uniform changes how helpful strangers are, and why?
