RESEARCH METHODS

• Experimental method involves manipulating an independent variable (IV) to affect the dependent variable (DV), which is then measured and stated in results
• Experiments can be field, laboratory, quasi, or natural
• Aims in research are general statements made by the researcher about what they plan to investigate
• Hypotheses are precise statements that clearly state the relationship between the variables being investigated, and can be directional or non-directional
• Independent variables are manipulated by the researcher to affect the dependent variable, which is then measured
• Operationalisation refers to clearly defining variables in terms of how they are measured
• Extraneous variables are variables that affect the dependent variable (DV) and do not vary systematically with the independent variable (IV)
• Confounding variables are variables other than the IV that have an effect on the DV and change systematically with the IV
• Demand characteristics are cues from the researcher or research situation that may influence participant behavior
• Investigator effects refer to unwanted influences from the researcher's behavior on the measured results
• Randomisation and standardisation are used to minimize the effects of extraneous or confounding variables in research
• Sampling methods include opportunity sampling and random sampling
• Opportunity sampling involves recruiting conveniently available participants, while random sampling gives all members of the population an equal chance of selection
• Laboratory experiments offer high control but low ecological validity, while field experiments provide natural behaviors but may lack control over extraneous variables
• Quasi experiments have controlled conditions but may lack random allocation of participants, affecting internal validity
• Natural experiments deal with real-life issues but may lack replicability and generalizability due to the rarity of natural events
• Random sampling:

  • All members of the population have equal chances of being selected
  • Each member is assigned a number, then a random number table, generator, or lottery method is used to choose a partner
  • No researcher bias as the researcher has no influence on who is picked
  • Time-consuming as it requires a list of population members (sampling frame) and contacting them
• Volunteer bias:

  • Participants can refuse to take part, leading to an unrepresentative sample
  • Participants may not take the study seriously if motivated by factors like money
• Systematic sampling:

  • A predetermined system selects every nth member from the sampling frame consistently
  • Avoids researcher bias and is usually fairly representative of the population
  • Not truly unbiased unless a random number generator is used to start the systematic sample
• Stratified sampling:

  • Sample composition reflects varying proportions of people in specific subgroups (strata) within the population
  • No researcher bias as selection within each stratum is done randomly
  • Produces representative data due to proportional strata, enabling generalization
• Volunteer sampling:

  • Involves self-selection where participants offer to take part
  • Quick access to willing participants but can lead to volunteer bias and affect generalizability
• Experimental Design:

  • Independent groups design: participants perform in one condition of the independent variable (IV), eliminating order effects and demand characteristics
• Experimental Design:

  • Repeated measures: the same participants take part in all conditions of the IV, eliminating participant variables but presenting order effects
• Experimental Design:

  • Matched pairs: pairs of participants are matched on a variable affecting the dependent variable (DV), one member does one condition and the other does another, reducing order effects but requiring a large pool of potential participants
• Pilot Studies:

  • Small-scale versions of investigations done before the real study to identify potential problems and modify procedures
• Single-blind and Double-blind Procedures:

  • Single-blind: researchers do not inform participants if they receive a test or control treatment to avoid bias
  • Double-blind: neither participants nor the experimenter know who receives a treatment to prevent bias
• Observational Techniques:

  • Naturalistic observation: watching and recording behavior in a natural setting for high ecological validity but can lead to awareness and replication difficulties
• Observational Techniques:

  • Controlled observation: watching and recording behavior in a structured environment for more control over variables but low ecological validity
• Observational Techniques:

  • Overt observation: participants are aware they are being watched, ethically acceptable but likely to record unnatural behavior
• Observational Techniques:

  • Covert observation: participants are unaware they are being watched, recording natural behavior but raises ethical issues
• Observational Techniques:

  • Participant observation: the researcher is part of the group being observed, providing insight but risking behavior changes
• Observational Techniques:

  • Non-participant observation: the researcher observes from a distance, increasing objectivity but open to observer bias
• Observational Designs:

  • Unstructured observation: continuous recording for richness of detail but higher risk of observer bias
• Observational Designs:

  • Structured observation: quantifying observations using predetermined behaviors and sampling methods for systematic data collection but less depth of detail
• Observational Designs:

  • Behavioral categories: breaking down target behaviors into observable components for precise observation and measurement
• Observational Designs:

  • Inter observer reliability: checking the agreement among researchers conducting the study to ensure unbiased reports
• Observational Designs:

  • Inter observer reliability score above 80% indicates high reliability
• When forming a behavioral categories list, it's crucial to ensure behaviors don't overlap with other similar behaviors, and they should be clearly operationalized
• Structured interviews involve different types of sampling methods:

  • Time sampling: records behavior within a pre-established timeframe, reducing the number of observations needed but may be unrepresentative
  • Event sampling: counts the number of times a specific behavior occurs, useful for infrequent behaviors but may overlook important details or have counting errors
• Correlations are used to investigate associations between two variables, measured but not manipulated like in experiments