Unit 1 Research Methods

• LAB EXPERIMENT - Scientifically controlled environment, untypical for natural behaviour, researchers invite participants, cause and effect of IV + DV by high exercise of control over eliminating extraneous variables, best for behaviour unaffected by environment (memory test)

  
  

  P: Cause and effect between IV & DV for (internal) validity.

  C: Invited participants may have demand characteristics

  C: "Reductionist" - oversimplifying something complex; ignoring other variables that could be involved

  
• FIELD EXPERIMENTS - Natural environment; unpredictable so not as high control of extraneous variable, IV manipulated by researcher and DV measured, overt or covert (participants may(not) be aware), best for behaviour in natural environment i.e conformity, obedience

  P: High ecological validity - generalisable to target population, less demand characteristics.

  C: Unpredictable so not all extraneous variables controlled, may bias/confound results. Confounding variables e.g age and sample bias without random allocation.

  C: Breaches ethics: informed consent, right to withdraw.

  
• NATURAL EXPERIMENT - IRL environment, IV + DV without manipulating IV as it naturally happens, opportunist unique and naturally occurring solutions otherwise hard or unethical to set up e.g: smoking banned, investigators reported 60% drop in heart attacks; no scientists can directly change something, but it rather naturally influences an effect in IV)

  P: real and naturalistic enviorment

  C: No random allocation to conditions of IV; participant variables e.g personality confounds findings

  C: Extraneous variables difficult to control

  NOTE: covert or overt

  
• Types of Validity:

  (Internal) Validity -  the measures were what they were supposed to measure.

  External Validity - the findings are generalisable to the target population - look at sampling method: can’t generalise sample until you repeat. Also, stratified sampling is good because represents even small percentage of population.

  Validity in Design - how participants allocated to groups can lead to validity issues.

  NOTE: If sample error (differing in intended qualities from target population) occurs, the findings of research are untrue of target population, affecting external validity.