Research Methods

• Experimental Methods


  Involves manipulation of variables to determine cause and effect, participants randomly allocated, all procedures standardised, so the same for all participants
• Aim

  General statement of what the researcher intends to investigate, purpose of the study, aims are developed from theories
• Hypothesis


  Clear, precise, testable statement that states the relationship between the variables being investigated
• Types of Hypothesis

  • Directional (one tailed, clear difference anticipated between 2 conditions, contains words like more/less, higher/lower)
  • Non-Directional (2 tailed, simply states there will be a difference between conditions but the difference isn't specified)
• Directional Hypothesis

  Used when findings from previous research suggest a particular outcome
• Non-Directional Hypothesis


  Used when there's no previous research or research contradicts
• Variables


  Anything that can vary or change within an investigation
• Independent Variable (IV)


  Factor manipulated by researcher, its effect on DV is measured, can be changed by researcher
• Dependent Variable (DV)


  The factor measured by researcher, effects on DV caused by manipulation of IV
• Operationalisation

  Clearly defining variables into measurable factors
• Standardisation

  Using exactly the same formalised procedures and instructions for all participants in a research study
• Extraneous Variables


  Variables besides IV that could affect DV and need to be controlled so they don't affect results
• Types of Extraneous Variables

  • Participant Variables (age, intelligence)
  • Situational Variables (experimental settings, weather, noise levels, area)
  • Experimenter Variables (changes in appearance, personality or conduct of research)
• Experimental Designs

  • Independent Groups Design (IGD)
  • Repeated Measures Design (RMD)
  • Matched Pairs Design (MPD)
• Independent Groups Design (IGD)


  • Uses different participants in each of the experimental conditions so each participant only does 1 condition (experimental/control), different participants are tested against each other
  • No order effect, as different participants so no effect whereby order in which conditions are done may affect outcome
  • Less economical, as 2x participants needed
• Repeated Measures Design (RMD)

  • Each participant tested in all conditions of an experiment, participants are tested against themselves
  • Because the same people are being measured in both conditions there's no individual differences between groups which may have caused a difference otherwise
  • Order effects, with RMD participants do all conditions and order in which they do these conditions can affect the results, participants may perform worse in 2nd condition due to fatigue/boredom
  • Chance of Demand Characteristics as if they participate in all conditions more likely they may guess the purpose of study and act a certain way due to this
• Matched Pairs Design (MPD)

  • Different but similar participants are used in each condition, participants are matched on characteristics important for particular study
  • Order effects are less likely
  • Participants aren't able to be matched exactly, even if identical twins are used there will be some difference between them, so individual differences remain
  • Matching participants may be time consuming and expensive, especially if pre-test is required, less economical than most other designs
• Types of Experiments

  • Lab Experiments
  • Quasi Experiments
  • Field Experiments
  • Natural Experiments
• Lab Experiments

  • High control environment, standardised procedure, randomly allocated experimental groups
  • High degree of control, so greater accuracy
  • Easy to replicate, so others can repeat and check results
  • Cause and Effect, other variables are controlled so you can establish cause and effect of manipulated variables
  • Low ecological validity, done in artificial scenarios so unlike real life, difficult to generalise results
  • Demand characteristics, participants aware they're being tested so may display demand characteristics, act unnaturally
• Quasi Experiments

  • The IV occurs naturally, e.g. sex, so study on effects of sex on say learning the 2 conditions of gender occur naturally
  • Shares the same strengths as lab experiments due to its controlled conditions
  • Can't randomly allocate participants to conditions, may have confounding variables
• Field Experiments

  • Conducted in naturalistic, real world, environmental setting, IV manipulated and experimenter tries to control as many other variables as possible
• Natural Experiments

  • In natural experiments IV changes naturally
  • High ecological validity, real world environment, relates to everyday behaviour so can be generalised to other settings
  • No Demand Characteristics, participants unaware of the experiment so they act naturally, no demand characteristics
  • Less control, more difficult to control extraneous variables so more difficult to establish cause and effect
  • Ethics, if participants aren't aware they're in an experiment there's lack of informed consent
• Sampling

  Researchers generally don't have means to test whole population so sampling is used, part of the population, who are ideally representative of population it was taken from
• Random Sampling

  Each member of population has equal chance of being selected, one way is place all names in container and pick the amount needed, another way is list of all members of population obtained, all names on list are assigned a number, sample generated through lottery method
• Random Sampling


  • Unbiased selection, increases the chances of getting an unbiased and representative sample
  • Generalisability, because the sample should be fairly representative results will be generalisable to the target population
  • Unrepresentative, just because selection is unbiased it doesn't guarantee an unbiased selection as all females and no males could be selected through this method
  • Difficult and time consuming to conduct
• Systematic Sampling

  Involves taking every nth person from list to create a sample, involves calculating the size of the population and the assessing what size the sample needs to be to work out sampling interval
• Systematic Sampling

  • Unbiased selection, no bias in selecting therefore increasing chance of unbiased representative sample
  • Generalisability, results are representative of population therefore you can generalise results
  • May not be representative, unbiased selection doesn't guarantee unbiased sample e.g. all females could be selected through this way
• Stratified Sampling


  Involves dividing population into characteristics important for research e.g. age, class etc then the population is randomly sampled within each category
• Stratified Sampling

  • Representative, selection occurs from representative sub-groups within population so should be fairly representative
  • Unbiased, it's performed upon sub-group of population selection is unbiased
  • Time-consuming, dividing of population into categories and then randomly selecting can be time consuming
• Opportunity Sampling


  Selecting those who are available and willing to take part e.g. asking people in the street who are passing
• Opportunity Sampling

  • Convenient, easy to form as they use people who are readily available
  • Unrepresentative, sample likely to be biased as certain types of participants won't take part therefore findings can't be generalised
  • Self-selected, as participants have the option to decline to take part those taking part are self selected
• Volunteer Sampling


  People volunteer to participate, so they select themselves as participants, often through replying to ads
• Volunteer Sampling

  • Ease of formation, requires little effort from researchers as participants volunteer
  • Less chance of screw you effect as the volunteers participate so less chance they will deliberately try to sabotage study
  • Unrepresentative, sample will be biased as volunteers are certain type of people and so you can't generalise results
  • Demand characteristics, if they volunteer they may be eager to please the researcher and therefore give answers they think researcher wants to hear
• Ethical Issues


  Researchers assess and act on ethical considerations involved in a research before it's conducted, main consideration: health and dignity of participants is protected
• Ethical Considerations

  • Informed consent
  • Avoidance of deception
  • Adequate briefing/debriefing
  • Protection of participants
  • Right to withdraw
  • Confidentiality
  • Observational research
  • Incentives to take part
• Informed Consent

  Investigators should give participants sufficient details of investigation so they can decide if they want to participate, parental consent should be obtained if child is under 16, can't get informed consent from those under influence of alcohol, drugs or if they're mentally unfit to give consent
• Avoidance of Deception


  Withholding of information or misleading participants is unacceptable if participants are likely to object once debriefed, participants shouldn't be deliberately misled without scientific or medical justification, can't gain informed consent if deception occurs, if deception is used, participants must be told immediately afterwards and given opportunity to withhold their data from study
• Pilot Studies

  Small-scale practice investigations carried out prior to research to identify potential problems with design, method or analysis so they can be fixed
• Control Groups & Conditions

  Group receiving the placebo sets the baseline and is there to act as a purpose of comparison, if a change in behaviour of experimental group is greater than that of the control group, researcher is able to conclude cause and effect was the IV
• Demand Characteristics

  Research involves interaction between participant and researcher, the interactions can affect findings, several features of research studies that enable participants to guess what study is about, this can cause demand characteristics like please-u, screw you effect, acting unnaturally due to nervousness, fear of evaluation or social desirability bias