Research Procedures (AQA GCSE Psychology): Revision Note

Standardisation

• Standardisation means keeping the procedure identical for every participant and every condition (apart from the IV itself), including:

  • the instructions given to participants, written out word-for-word in advance

    • This stops what is said acting as an EV, e.g. telling one participant to "do their best" but not saying this to another

  • the participant's right to withdraw. Participants should be told they can leave the study at any time

  • the briefing and debriefing

  • setting – participants tested in the same environment

    • This is easier to control in a laboratory than, say, testing people in different coffee shops

  • timings – each condition runs for the same length of time (unless a time delay is the IV)

  • materials – identical, or equivalent in difficulty where the design requires different versions

    • E.g. in a repeated measures design, two word lists matched for difficulty so participants don't benefit from seeing the same list twice)

• Standardisation means the study can be replicated exactly

  • If replications produce similar results, the findings are reliable (consistent)

Yuki et al.'s (2007) emoticon study used a standardised procedure.

Randomisation

• Randomisation means using chance to decide key aspects of the design, rather than the researcher deciding, as this reduces bias, e.g.:

  • random allocation of participants to conditions, e.g. drawing names from a hat

  • deciding the order of items in a word list randomly, so easy or related words don't cluster together and unfairly help recall

• A limitation is that chance can still produce unbalanced outcomes

  • E.g. one condition ending up mostly female, or a word list with several words starting with the same letter

• In a repeated measures design participants do all conditions, so random allocation isn't possible

• Instead counterbalancing can be used as it controls order effects (practice, fatigue)

• Counterbalancing is where half the participants do A then B, the other half B then A, so order effects are balanced out across the two conditions

How to control extraneous variables

• Extraneous variables are usually controlled by keeping them constant across all conditions, so they affect every participant equally and cannot explain any difference in the DV, for example:

  • time of day: run all conditions at the same time, e.g. midday

  • temperature: keep the experimental space at a constant temperature

  • mood: treat all participants in the same friendly, neutral way

• If an extraneous variable is not controlled and it varies systematically with the IV, it becomes a confounding variable

  • e.g. all Condition A participants tested in the morning, all Condition B tested in the evening, when time of day is not the IV

• A confounding variable means the researcher can no longer tell whether it was the IV or the confound that caused the change in the DV, therefore the study can no longer demonstrate cause and effect

Evaluation of procedures used in research

Strengths

• Controlling EVs increases confidence that the IV caused the change in the DV (internal validity)

• Standardised procedures also allow the study to be replicated; if repeated studies produce similar results, this shows the findings are reliable

• Being able to replicate a study lets other researchers keep testing the theory, e.g. Asch's conformity study has been replicated many times in different countries

Weaknesses

• Where there is high control there tends to be low validity, particularly ecological validity

• The artificial context and tasks involved in high-control research could give rise to demand characteristics which is another threat to validity