Finding Extraneous Variables and Controls
- In an apply question you first hunt for extraneous variables in a study, then suggest a control for each.
- Start by finding anything, other than the IV, that could affect the DV.
- Then propose a control that removes or reduces each one.
- Tie every control you name to a specific problem in the study.
- Decide first whether the problem is about the setting or about the people.
A Worked Example
- A researcher asks whether an energy drink speeds up reaction time.
- One group is tested first thing on Monday morning; the other straight after PE on Friday afternoon.
- Situational variable: the time of day and the day of the week differ between the groups.
- Participant variable: the Friday group is tired after PE, which could slow their responses.
- Controls: test both groups at the same time of day in the same room, and allocate people to conditions at random.
- With the setting held constant, any difference in reaction time now points to the drink.
- Leave the PE fatigue uncontrolled and it gives an alternative explanation for the slow times.
Choosing the Right Control
- For a situational variable, standardise the setting so it is the same for everyone.
- For a participant variable, use random allocation to spread differences across the conditions.
- For order effects in a repeated measures study, use counterbalancing.
- Do not suggest a control without saying which problem it fixes.
- Do not name the IV itself as an extraneous variable.
A Routine for Apply Questions
- List what differs between the conditions in the scenario apart from the IV.
- Label each as a situational or participant variable.
- Match a control to each and say which problem it fixes; the link to that problem scores separately from naming the control.
- What two steps does an apply question on extraneous variables need?
- In a study where one group is tested after lunch and one before, name an extraneous variable and a control.
- Which control fixes a participant variable?
- Why must you link a control to a specific problem?