Presentation on theme: "Occupational coding: principles, practice and problems A workshop within the ESRC Research Methods Programme Peter Elias Institute for Employment Research."— Presentation transcript:
Occupational coding: principles, practice and problems A workshop within the ESRC Research Methods Programme Peter Elias Institute for Employment Research University of Warwick Royal Statistical Society, London June 2004
Principles of occupational coding Aim: to categorise an occupational description within an occupational classification Rationale: to reduce a complex set of information to a manageable set of categories which reflect the conceptual basis of the classification
Some definitions What is the unit to be classified? A job – a set of tasks/duties executed, or designed to be executed, by one person What is an occupational description? Account of the main tasks and duties performed in a set of jobs which are characterised by a high degree of similarity
What do we mean by similarity? Similar in terms of the underlying conceptual basis of the classification What is the conceptual basis This relates to the purpose for which the classification is designed (e.g. to measure skill, social positions, safety of the work environment, etc.) Some definitions (contd.)
What is an occupational classification? A set of categories which reflect the conceptual basis of the classification and within which all units can be appropriately classified Usually described via a Structure, an Index and a set of Rules Some definitions (contd.)
Coding practice Self classification Present the informant with a set of occupational categories – they select most appropriate for the occupation they wish to classify Coder classification All the requisite information is passed to a coder. Coding is achieved by comparing occupational information with the index to the classification, applying the coding rules to obtain the most appropriate code for the occupational category
Self-classification Consistent coding requires all informants to have the same understanding of the nature of the classification (or, at least, for errors in coding to be normally distributed)
SOC2000 Major groups 1 Managers and senior officials 2 Professional occupations 3 Associate professional and technical occupations 4 Administrative and secretarial occupations 5 Skilled trades occupations 6 Personal service occupations 7 Sales and customer service occupations 8 Process, plant and machine operatives 9 Elementary occupations Can you classify your job to SOC2000?
Coder classification Collect required information: Labour Force Survey What was your (main) job (in the week ending Sunday [date])? What did you mainly do in your job? Millennium Cohort (First survey) What is your (main) job? What do you mainly do in your job?
"name": "Coder classification Collect required information: Labour Force Survey What was your (main) job (in the week ending Sunday [date]).",
"description": "What did you mainly do in your job. Millennium Cohort (First survey) What is your (main) job. What do you mainly do in your job.
Compare information collected with index and apply coding rules - reverse word order? - default coding rules? -see notes? - obtain code Coder classification (contd.)
What problems can arise? Cannot always collect the information needed for detailed and accurate coding The index and/or the rules may not yield an appropriate code The process is complex, tedious and error prone
How do we resolve these problems? Give explicit instructions to informant, with examples Ask the right questions Update the index and the rules on a regular basis in the light of experience Coder training Interviewer coding Update the classification
Updating the Classification 198019902000 Classification of Occupations 1980 (CO80) 1990 Standard Occupational Classification (SOC90) 2000 Standard Occupational Classification (SOC2000) CODOT and the Key list of Occupations for Statistical Purposes (KOS) HE First destination statistics (SOC(FDS)) HE Destination of Leavers Survey (SOC(DLHE))
The classification of ICT occupations (1946-1972) ED526 (1946-72) Systems analyst/programmer (computers) Programmer (computers) Hollerith machine operator
CODOT (1973-90) Office manager (records) Data processing manager Systems analyst Computer programmer Data processing machine operator Computer operator Sorting machine operator Key punch operator The classification of ICT occupations (1973-1990)
SOC90 (1991 – 2000) Computer systems and data processing managers Software engineers Computer analysts/programmers Filing, computer and records clerks Typists and word processor operators Computer and data processing machine operators Computer engineers; installation and maintenance The classification of ICT occupations (1991-2000)
SOC2000 (2001 – 2010) ICT managers IT strategy and planning professionals Software engineers IT operations technicians IT user support technicians Database assistants/clerks Computer engineers; installation and maintenance The classification of ICT occupations (2001-2010)
How well can we monitor occupational change? Monitoring trends requires that we backcast occupational data when a new classification is introduced Different approaches to this: - recoding historical data (expensive) - use of converter matrices (inaccurate) - preserving text descriptions for future reprocessing
Classification converters? LFS Winter 96/97 and Spring 2000 dual coded – SOC90 and SOC2000 91 Census (E&W ½ % sample) component coded – CO80 and SOC90 1990 New Earning Survey – KOS and SOC90 For assistance, contact Occupation Information Unit
Growth of ICT-related occupations, England and Wales 1991- 2003
Good coding practice Code to the most recent version of the UK national occupational classification Adopt the wording of questions as developed by the Office for national statistics Create electronic records of text descriptions Train coders: promote understanding of conceptual basis of classification, nature of rules, use of index Where possible, dual code to previous occupational classification Conduct consistency checks