Sociology 2267A/B Lecture Notes - Lecture 3: Juvenile Delinquency, Convergent Validity, Face Validity
Document Summary
Police (uniform crime reporting) most commonly reported by the media. Court statistics typically mirror police statistics, not standardized. Self-report surveys researchers dramatically changed the field because they bring good data. Individual sources aren"t as good, more than one source is best. Criticism: individual sources do not give an accurate portrayal of youth offending. If we were to repeat the data collection, would we get the same results assuming all things are equal: same scores should be achieved across time. Things that can affect reliability: personal bias (e. g. , memory/recall, external factors (e. g. , funding, district priorities) Measurements need to be stable, dependable, replicable, generalizable: across time test-retest reliability. Same results should be found if given to same group over time. Different results not stable: across instruments parallel test reliability. Different types of surveys getting at the same thing should tell the same thing. Reliability of the data: across items split-half reliability (internal consistency)