Official Statistics

Official Statistics

1. Practical

   • Access
     • Available to public
   • Subject matter
     • Study the past
   • Free source of huge amount of data
     • Some are compulsory (births/deaths)
   • Statistics created for a purpose independent of the sociologist
     • French state: No stats on ethnicity, race, religion
   • Large scale
     • Census = nationwide
     • General household survey = 23,000 people
     • BCS: 47,000

2. Theoretical

   • Validity
     • Definitions
       • May differ between state and sociologist
       • Change over time
     • Social construct
       • Cicourel: NOT social facts
       • Suicide
         • Douglas
     • Soft statistics
       • State agencies
       • 'Dark figure'
         • Racist incidents in schools: record all REPORTED
         • 90% public reports
         • BCS: Police report 60% crime
     • Hard statistics
       • Registration data (births/deaths)
         • Clear catagories
         • Legal requirement
     • Bias towards 'notable offencces'
   • Representativeness
     • Census completed by 95%
     • Official surveys are less
     • Response inconsitency
     • Careful sampling
   • Reliability
     • Standardised measuring instrument
     • Standard procedure
     • Human error
       • Filled out incorrectly
       • Coder's errors
     • Tombs: employee deaths due to h&s breaches missing from stats on violent crime
       • Murder/manslaughter
     • DW: Series of acts = 1 offence
   • Hypothesis testing
     • Patterns over time
       • Changes in law
     • Cause and effect
       • Durkheim's study of suicide
     • Possible mismatches between sets of statistics
   • Objective
   • (M/F) Ideological functions
     • Althusser: ISA
     • Conceal/distort reality
       • Capitalism
         • Changing definition of unemployment
           • Thatcher's Government
       • Patriarchy
         • Few stats on unpaid domestic labour
         • 2001: Class determined by 'HRP'