PSY 313 Lecture Notes - Lecture 3: Operational Definition, Confounding, Random Assignment
Document Summary
Inter-rater reliability: correlation between different raters; same observations, different raters. On a test: gets a score of 5 for questions 1-5 and a 4 for questions 6-10. Observed score/measure = true value +/- error (ex. exam score = knowledge +/- stress) Random error: due to instrument or person being measured. Intrinsic noise: drink a little bit of water between 2 measurements, you may weight more: measurement/observer error: reading from a scale. Cancels out with repeated measures because it"s random. Reliability is necessary for validity: external generalization to diff. times, places, people (are the results global) Population are the results generalizable to other people/groups. Ecological generalizable to other environmental conditions/settings: internal alternate causes ruled out. Confounds are a threat to internal validity (prevents a causal conclusion) because they explain the results but are not the iv. Extraneous variable (ev): any variable, other than the iv, that affects the.