TOWNSEND PART 3

TOWNSEND PART 3

1. From concepts to indicators

   • From an abstract and complex of poverty (socially constructed) to a set of indicators (measurement)
   • De Vaus: "The process of moving to the point where we can develop questionnaire items to tap the concept... the descending ladder of abstraction"

2. From theory to measurement

   • Concept specification
   • Operationalization
     • Variables to indicators
   • Measurement
     • Criteria for validity and reliability
       • Very important

3. How did Townsend identify his indicators?

   • Assumed a uniform standard of living
     • This has been criticzed
     • 60 - safe conclusion?
   • Based on expert opinion
   • No exhaustive study of lifestyles
   • Assumes national style of living
     • Chose indicators to cover "all the major areas of personal, household and social life"
     • Uses 'lack of something' as deprivation

4. Evaluating measures

   • Is poverty being measure consistently? Reliability
     • End with same answers/results on repeated occasions?
       • Multiple indicators often more reliable
         • Shouldn't matter who asked the initial question
   • Did Townsend measure what he intended to measure? Validity
     • Criterion Validity
       • Confident measurement = consistent with other well-established measures of same concept
         • Townsend correlates deprivation indicators with low income and other measures i.e. state support
           • Income measurement valid - most people deprived of list items = state support or low income
     • Content validity
       • individual indicates may not be sufficient on there own - but together they measure different aspects of the concept
         • Townsend: no single item on its own can measure poverty
           • High score on index - 'highly suggestive'
             • Without one item - not a measure i.e. vegetarianism
     • Construct validity
       • Does the indicator really tell us about what we are trying to measure?
       • How well does the measure conform with theoretical expectations
       • Theory -> indicator design -> theory

5. De Vaus 2002: What is a survey

   • Methodology distinguishable from the intrusnments or techniques used to generate data
     • E.g. questionnaire, semi-structures interviews, observation
       • Very basic
   • Survery distinguishable by form of data; method of analysis

6. De Vaus 2002: What will a survey do

   • Collection of systematic data that allows for systematic comparison between cases on the same characteristics (e.g. age and gender)
   • Find correlation
   • "An investigation where systematic measurements over a series of cases yielding a rectangle of data.... The variables in the matrix are analysed to see if they show any patterns"
   • BUT "only by understanding intentions, ideas... that motivate people can you really know anything" Weber's verstehen

7. Survey Magic - sampling theory

   • if representative - can generalise from small sample to large population
     • Simple random sampling
       • Gallup's probability sampling of 5000 more accurate than Literary Digest's volunteer sample of 2 million
       • Element of chance = necessary
         • No selection bias
           • Small sample = general population

8. Survey critiques

   • Role of context
     • Practical problems
       • Non-response
         • social desirability

9. Measurement and statistics

   • Some things cannot be measured
     • Reduction of human behaviour to numbers
       • Infinetly complex world
         • Opaque: simply tools to implement various logical operations

10. De Vaus: interpretation is fundamental to a well-concieved surgery research and analysis (deductive = good)

11. Problems of standardisation

    • Understand questions?
      • Asked the same - T used few researchers

12. "Three lies: lies, damned lies and statistics" Mark Twain