11 12 2012 21 25 00 Research Methodology Independent variable causes of Dependent variable outcome what we are attempting to explain Correlation and causality Just because 2 items are correlated does NOT imply a causal connection Temporal Time Element Cross sectional date data collected at only 1 point in time Longitudinal data data collected at more than 1 point in time Levels of Analysis Macro societal nations states cities Micro individual Establishing causality 3 criteria Correlation necessary but insufficient Temporal ordering cause has to proceed affect Spuriousness must be eliminated outside elements being the actual cause Dark Figure of Crime Unknown amount of crime that happens 3 General Sources of Crime Statistics Official Reports Victim Surveys Self Reports Started in 1930 Uniform Crime Reports UCR FBI Crime in the US 18 000 law enforcement agencies reporting representing 95 of the population 8 crimes listed homicide rape robbery agg assault burglary theft MV theft arson index crimes results calculated as a rate of crimes population 100 000 rate includes the Supplemental Homicide Report advantages trend data consistent definition of crime limitations subject to political manipulation national figures can be broken down represent census hierarchy rule only takes into account most serious underestimates despite best efforts still differences in definitions ecological fallacy don t guess what is contributing to crime measure of police activity rather than actual crime National Crime Victimization Survey NCVS Started in 1972 Household survey conducted by census Bureau Members older than 12 surveyed by phone 3 year inclusion 6 month intervals uses sampling to produce national estimates advantages eliminates reporting report surveys by bounding disadvantages design reduces memory decay and telescoping seen in other self rape question redesign in 1992 produced higher estimates household survey excludes institutions like jails and businesses victimless crimes and homicide not included series victimizations produces undercounting only kids 12 and up excludes child abuse Total violent crime victimizations reported to the police crimes recorded by the police arrests for violent crime National Incident Based Reporting System NIBRS Started in 1984 Piloted by South Carolina approved by 1988 20 of US population was represented 16 covered by UCR also covered by NIBRS enhance quantity quality and timeliness of statistical data collected improve methodology used for compiling analyzing and publishing agencies that use NIBRS collect data and report them thorough 2 main goals by law enforcement crime data information on crime group A offenses street crimes serious white collar offenses group B offenses status crimes prostitution minor white collar crime advantages gets rid of hierarchal rule disadvantages still only reported crimes collects MORE data MORE crime MORE criminals uncertainties due to how young the system is compliment to others such as UCR Self Report Data Short Nye 1957 Violated a lot of assumptions people were making at the time about offenders remember victims don t always don t always report it juvenile delinquency National Youth Survey Longitudinal survey 1 cohort over time maps to UCR well in terms of crime rates advantages can see cause and outcomes over time covers victimless crimes gets a dark figure of crime gets at source not subject to manipulation politicization offender driven data disadvantages people lie can exaggerate or underestimate crimes limitations sampling re phone geographical false reporting focus more on minor crimes Easier to get people to admit to lesser crimes policy and practice must be sensitive to what research has de 11 12 2012 21 25 00 11 12 2012 21 25 00
View Full Document
Unlocking...