PSYCH 3B03 Lecture Notes - Lecture 10: Naturalistic Observation, External Validity, Selection Bias
Document Summary
The goal of research is:to determine cause and effect relationships. Random sampling isn"t always perfect in child psychopathology research because: Clinicians tend to sample children already in hospital/receiving treatment. People already being treated tend to have more severe symptoms; research then involves more of a bias towards severe symptoms. Selection bias: any error that can happen while random sampling that prevents sample from being generalizable to the population. Operational definition : an observable and measurable definition of the concept being studied. Example: aggression = the frequency that a child physically harms or threatens another person. Naturalistic observation: directly observing individuals in the real world without interference, recording how many behaviours occur based on the observational definition. Best measure of certain behaviours because you see how children act naturally. Descriptive (non- experimental): portrays a phenomenon of interest using observational methods. Tend to use observational methods; not controlling the situation.