Experimental Methods

• Laboratory experiments- controlled environment, manipulates independent variable and measures it's effect on dependent variable, extraneous variables controlled
• Strengths of lab experiments:

  • High internal validity- control of variables means cause and effect can be determined
  • Highly replicable- increases reliability of results
• Limitations of lab experiments:

  • Low generalisability- artificial settings means results can't be applied to other settings
  • Low ecological validity- due to low generalisability
  • Low mundane realism- tasks done in the setting won't be the same as normal tasks
  • Risk of demand characteristics- participants are aware they are being studied due to the setting
• Field experiments- natural setting, manipulates independent variables and measure the effect on dependent variables
• Strengths of field experiments:

  • Higher mundane realism- tasks are done in a real-life setting where they might occur normally
  • Higher ecological validity- done in a 'normal' setting, can be more generalisable
• Limitations of field experiments:

  • Less control over variables- can be more difficult to establish link between cause and effect
  • Ethical issues- participants may not consent to being a part of the study, their privacy could be jeopardised
• Natural experiments- independent variable is naturally occurring in the environment and not manipulated by the researcher, effect on dependent variable is measured
• Strengths of natural experiments:

  • Allows research into otherwise unethical topics
  • High ecological validity- real-life setting is more generalisable
• Limitations of natural experiments:

  • Low internal validity- independent variable isn't manipulated, so cause and effect can't be determined
  • Difficult to replicate- lower reliability
  • Participants can't be randomly allocated
• Quasi experiments- independent variable isn't manipulated by the researcher, it naturally exists in participants and it's effect on the dependent variable is measured
• Strengths of quasi experiments:

  • Controlled- increases replicability which also increases reliability
• Limitations of quasi experiments:

  • Low internal validity- independent variable isn't manipulated so cause and effect can't be determined
  • Participants aren't randomly allocated
  • Risk of confounding variables- lowers internal validity
• Repeated measures- the same pps take part in each condition of the experiment
• Strengths of repeated measures:

  • Less pps needed, lower risk of participant variables, less time consuming, easy to compare results between conditions (to same pps)

  Limitations of repeated measures:

  • Higher risk of demand characteristics (pps might change behaviour due to their impressions of what the study might be), order effects (more practise or fatigue) which could lead to skewed results or drop-outs which wastes resources
• Independent groups- different, randomly allocated participants take part in each experimental condition
• Strengths of independent groups:

  • Less risk of demand characteristics, no order effects because pps don't do both conditions

  Limitations of independent groups:

  • More pps are required, risk of participant variables, harder to compare results between conditions
• Matched pairs- pps are assessed on key variables relevant to the study, then matched into a pair with a pp in another condition based on the similar key variable
• Strengths of matched pairs:

  • Low risk of participant variables, no order effects, lower risk of demand characteristics, easy to compare results between conditions

  Limitations of matched pairs:

  • Time-consuming to match pps, hard to exactly match pps
• Counterbalancing- lowering the possibility of extraneous variables in a repeated measures design by splitting the pps group in half- one group will do condition 1 first then condition 2, the other group will do condition 2 first then condition 1, which will balance out any order effects
• Participant variables- individual differences between pps which might affect the results/ performance of an experiment