Research Methods

Research Methods

1. Experiments

   • Field
     • Involve manipulating the IV to see the effect on the DV
     • Takes place in a 'natural' environment
     • Lack of control over variables
     • Has ecological validity
   • Natural
     • Can be difficult to replicate
     • IV is not manipulated
   • Lab
     • Controlled conditions
     • Tests the effect of IV on the DV and controls as many EV's as possible
     • Involves the comparison of two or more conditions between which only one factor is changed
     • Doesn't reflect real life - lacks ecological validity
     • Can create artificial behaviour

2. Variables

   • IV
     • Independent variable
     • Manipulated by researcher
   • DV
     • Dependent variable
     • Measred by the researcher
   • EV
     • Extraneous variable
     • Other factors that may have an effect on the DV
       • eg. random errors, situational variables, participant variables, order effects, demand characteristics, experimenter effects
   • Operationalisation
     • Making a variable precise enough to be manipulated, measured or controlled

3. Hypothesis

   • A testable statement
   • Experimental/alternative hypothesis predicts there will be a difference or an expected relationship
   • Null hypothesis states that the results will not show a difference or relationship between the two variables
     • Any difference shown will be due to chance factors
   • Directional or non-directional
     • Directional (one tailed) - states the direction of the results
     • Non-directional (two tailed) - doesn't state which direction the results will go in

4. Ethics

   • Guidelines can be broken if it is justified
   • Consent - don't need consent in a field exp. but do in a lab exp.
   • Debrief - exp. should be discussed w/ participants afterwards and told the full extent of the study
   • Confidentiality - all info about participants should be confidential
   • Deception - should be avoided but sometimes it's necessary
   • Withdrawl - doesn't apply to obedience investigations
   • Protection of participants - participants should be safe from psychological and physical harm

5. Types of data

   • Qualitative
     • Consists of words
     • Strengths
       • Allows for more depth of analysis
       • High validity
     • Weaknesses
       • Difficult to make comparisons between groups or make a specific conclusion
       • Low reliability
   • Quantitative
     • Consists of numbers and quantities
     • Strengths
       • High reliability
       • Possible to analyse data in order to make comparisons between groups and make specific conclusions
     • Weaknesses
       • Gives a superficial view of the behaviour being studied
       • Low validity

6. Types of Interview

   • Structured
     • Has a pre-set order of questions , leaves little room for the researcher to follow up on answers
   • Unstructured
     • Questions are open and structure is flexible
   • Semi-structured
     • Has an order of questions to be answered but researcher has more freedom to follow up on answers

7. Sampling

   • Opportunity
     • Uses whoever is immediately available
     • Strengths - fast and efficient
     • Weaknesses - biased results
   • Volunteer
     • Uses whoever puts themselves forward
     • Strengths - less biased
     • Weaknesses - takes longer
   • Stratified
     • Rarely used
     • Sample reflects the target population
     • Strengths - very representative
     • Weaknesses - difficult to do
   • Random
     • Every no. of the target population has an equal chance of being selected
     • Strengths - v. representative
     • Weaknesses - takes a long time

8. Participant Designs

   • Independent Measures
     • Each group takes part in one experiment
     • Takes less time
     • No order effects
     • Large sample sizes
     • Individual differences
   • Repeated Measures
     • Each group takes part in all experiments (experiments repeated)
     • Takes longer
     • Order effects can occur
     • Smaller samples
     • No individual differences
   • Matched Pairs
     • Similar participants matched together in pairs
     • Individual differences controlled
     • No order effects
     • Takes longer