Research Methods

• INDEPENDENT


  • Groups
  • One condition
  • OC IV
  • No order
  • Less demand
  • Characteristic
  • Only aware of one condition
  • Cumulates
  • No counterbalancing
  • Over repeated measures
  • Different types
  • 0+ people
  • Changes to DV
  • Order effects
  • Random allocation
  • Equal chance
• REPEATED MEASURES


  • Same participants take part in all conditions
  • Not time consuming
  • May change findings
  • Conditions in one order, other half do opposite
• MATCHED PAIRS


  • Pairs of participants matched on same variable (age, diet)
  • One member does one condition, the other does the other
  • Less demand characteristics
  • Time consuming
  • Expensive
  • Large pool of participants needed
  • Difficult to gather variables
• NATURALISTIC OBSERVATION


  • Natural setting
  • High ecological validity
  • Uncontrolled extraneous/confounding variables
  • Unnatural behaviour
  • Low ecological validity
• CONTROLLED OBSERVATION


  • Structured environment, lab
  • High ecological validity
  • Can focus on specific aspects
  • Lower ecological validity if participants know they're being observed
  • Unnatural behaviour
  • Low ecological validity
• OVERT OBSERVATION

  • Participants know they're being watched
  • Unnatural behaviour
  • Demand characteristics
  • Ethically acceptable
• COVERT OBSERVATION


  • Participants unaware they're being watched
  • Natural behaviour
  • Unethical
  • Lower internal validity
  • No consent
• PARTICIPANT OBSERVATION


  • Researcher part of group being observed
  • More insightful
  • Lose objectivity
  • Researcher may identify too strongly
• NON-PARTICIPANT OBSERVATION


  • Researcher observes from a distance
  • Researcher can be more objective
  • Not part of study
  • Identify less
  • Observer bias
  • Stereotypes known to observer
  • May lose insight and ideas
• OPPORTUNITY SAMPLING


  • Participants must be available
  • Saves time
  • Less costly
  • Lacks generalisability
  • Researcher bias
• RANDOM SAMPLING

  • All participants have equal chances of being selected
  • No researcher bias
  • Time consuming
  • Volunteer bias
  • Unrepresentative
• SYSTEMATIC SAMPLING


  • Every nth member
  • No researcher bias
  • Representative of specific set of people
• STRATIFIED SAMPLING

  • Reflects varying proportions of people in a population
  • No researcher bias
  • Time consuming
  • Representative
• VOLUNTEER SAMPLING


  • Offer to partake
  • Not time consuming
  • Volunteer bias
  • Specific people attracted
• LAB EXPERIMENT


  • Highly controlled
  • Manipulates IV to see effects
  • Can be replicated
  • High internal validity
  • Low ecological validity
• FIELD EXPERIMENT


  • Natural setting
  • IV manipulated to see effects
  • High ecological validity
  • Lower internal validity
• NATURAL EXPERIMENT


  • IV not brought about by the researcher
  • High external validity
  • Naturally occurring IV (sex, age)
  • Lower internal validity
• QUASI-EXPERIMENT


  • IV not determined by researcher
  • Conditions naturally occurring
  • Higher internal validity
  • Lower ecological validity
  • Unethical
  • Invasion of privacy
  • Loss of control over extraneous variables
  • Difficult to replicate
• QUALITATIVE DATA


  • Written data
  • Independent
  • Rich in data
  • Allows participants to further develop opinions
  • External data
  • Meaningful data
  • Difficult to analyse
  • Difficult to compare
  • Researcher bias
• QUANTITATIVE DATA


  • Numerical data
  • Can be statistically analysed
  • Lacks detail and depth
  • Can be converted to graphs
  • Easy to compare
  • No meaningful insight
  • No first-hand information
  • No participant opinions
• PRIMARY DATA


  • Targets exact information
  • Fits aims and objectives
  • Requires time and effort
  • Expensive
• SECONDARY DATA


  • Second-hand information
  • Minimal effort to collect
  • Outdated/incomplete
  • May not be reliable or valid
• META-ANALYSIS


  • Combines results from many different studies
  • Replicable
  • Lots of data
  • Bias - researcher may not publish all data intentionally or false representation
• Inter-Observer Reliability


  Total number of agreements / Total number of observations
• UNSTRUCTURED CONTINUOUS RECORDING


  • Researcher writes down everything they observe
  • More richness and independent data
  • Lower reliability
  • Qualitative data
  • More difficult to record
• STRUCTURED OBSERVATION


  • Researcher quantifies what they are observing
  • Predetermined list of behaviours to observe
  • Easier/more systematic
  • Quantitative data
  • Lower risk of observer bias
  • Easier to analyse and compare
• TIME-SAMPLING


  • Recording behaviour in a pre-established time-frame before study
  • Reduces number of observations and analysis
  • Greater risk of observer bias
  • Can only record what they see/may leave out things
  • Less time-consuming
  • Good for infrequent behaviours
  • Lower inter-observer reliability
• EVENT-SAMPLING


  • Counting number of times a particular behaviour is carried out
  • Representitive
  • Predetermined bias
  • Subjective
  • Small data collections
  • Unrepresentitive of study
  • If behaviour is complex, may be overlooked
  • Frequent-counting errors
• Correlation


  • Can be used as starting points to assess potential relationships between co-variables before conducting experiment
  • Quick and economical
  • Uses secondary data
  • Less time consuming
  • Difficult to establish cause and effect
  • Chance or another variable
  • Researcher may be unaware
  • May be misused and misinterpreted by public and media
• An alternative hypothesis predicts what would happen if the null hypothesis was false, i.e., if there is a significant difference/effect
• The null hypothesis states that there will be no difference or effect on the dependent variable due to manipulation of independent variable
• A research hypothesis is an educated guess about the relationship between two variables