problems with official statistics
- problem with definitions e.g. unemployment rate
- problems with reporting and recording e.g. crime
- reflect political agendas
- often collected by researchers with little experience of using existing data sets
why is existing data a good thing?
- readily available=saves time and money
- better coverage and scope than i could collect
- collected by professional gives credibility
where can i find secondary data?
- office of National Statistics
- gap minder (bubble charts)
- UK data archive e.g. citizenship survey, English housing survey
population data defintion
every case you want to find out about e.g. census
administrative data definition
data that was collected without the intention of research e.g. education, marriages
problem with administrative data
may be confidential at an individual level
sample data definition
selection of cases from the population, ideally representative of that population
individual level data definition and example
data available on each case you are interested in e.g. marks each student received in a test . gives the maximum level of detail
aggregate data defintion and example
summary of the characteristics of individual data e.g. the scores a year group received in their gcse’s exams, 25 got an A, 6 got B etc
when is population data available (3)
1) only interested in a small number of cases that are easy to contact e.g. physical students at UOL
2) relevant data has been collected in some kind of census
3) you have access to administrative data (e.g. for health or education)
higher levels of aggregate data
cases are put into a smaller number of larger groups
lower levels of aggregate data
cases are put into a larger number of smaller groups
the more you summarise your data..
the higher the levels of aggregation
cross-sectional data defintion and example
snapshot of a person/group at a particular point in time. e.g. A level results in 2011
repeated cross-sectional data definition and example
similar to cross sectional data but carry out the same study at a different point in time, showing historical change . e.g. a level results between 2001-2011
longitudinal data defintion and example
data measured over time showing biographical change e.g. academic attainment of a student over the course of their undergraduate degree
uses of secondary data
- to formulate research questions
- contextualise data
- principle data source
- comparative data source
- help with sample design
limitations of secondary data
- might not be specific to your investigation
- don’t know what conditions it was collected in
- not always compatible/comparable
- data is often aggregated (=no individual level, but this can be a good thing)
