Holding Everything but the IV Constant
- Controls are the steps you take to reduce or hold constant the effect of extraneous variables.
- They keep everything except the IV the same across your conditions.
- Good controls let you be far more confident that the IV, and nothing else, caused the change in the DV.
- Each control tackles a particular kind of extraneous variable.
- Your goal is a fair comparison in which only the IV differs.
The Main Controls
Standardisation
Keeping the procedure, instructions and conditions exactly the same for every participant so that everyone is treated in the same way.
- Standardised procedures
- Keep the steps, instructions and materials identical for every participant.
- Randomisation
- Use chance to set the order of tasks or materials, so no bias creeps in.
- Random allocation
- Use chance to put participants into conditions, spreading participant variables evenly.
- Counterbalancing
- Half the participants do the conditions in one order and half in the other, cancelling out practice or fatigue effects.
- Standardise the setting so both groups sit the test in the same cool, quiet room rather than one lot in a stuffy school hall.
- Random allocation stops one class ending up with all the strongest readers.
Matching a Control to the Problem
Randomisation
Using chance, such as drawing names or a random number list, to decide things like the order of tasks or how participants are allocated, reducing the researcher's influence.
- For participant variables, use random allocation.
- For order effects like practice or fatigue, use counterbalancing.
- For situational variables, use standardised procedures to keep the setting the same.
- Do not just list controls; match each one to the problem it solves.
- Do not confuse randomisation of materials with random allocation of people to conditions.
Choosing Controls for a Study
- Name the exact extraneous variable in the study (situational, participant or order effect).
- Choose the matching control: standardisation for situational, random allocation for participant variables, counterbalancing for order effects.
- Explain how that control holds it constant; a control named without saying how it works only half-scores.
- What is the purpose of a control in an experiment?
- What does random allocation control, and how?
- What is standardisation, and why does it matter?
- Which control would you use for practice effects, and why?