Statistics Canada
description
Transcript of Statistics Canada
THE ETHNIC THE ETHNIC DIVERSITY SURVEYDIVERSITY SURVEY
(EDS)(EDS)
Content and Data Availability
Kelly [email protected]
416-952-1919Statistics Canada
June 21, 2005
Statistics Canada
Statistique Canada
Canadian Heritage
Patrimoine canadien
Outline of presentation
• Overview of the Ethnic Diversity Survey
– objectives
– survey design
– content
• Review of Ethnic Diversity Survey products
• Advice on use of data on the EDS Analytical File
• Description of the EDS Public Use Microdata File
Ethnic Diversity Survey objectives
To provide information on the ethnic and cultural backgrounds of people in Canada and how these backgrounds relate to their lives today
To provide information to better understand how Canadians of different ethnic backgrounds interpret and report their ethnicity
Survey funded jointly by Statistics Canada and the Department of Canadian Heritage
Target population & sample design
Non-Aboriginal individuals aged 15 years and older living in private dwellings in Canada’s ten provinces
57,200 persons selected to be interviewed between April and August 2002
Two-phase stratified sampling design based on responses to the 2001 Census ethnic origin, birthplace & birthplace of parents questions; 15 strata were created
Reference period & data collection
Computer Assisted Telephone interviews April to August 2002
Average length of interview: 35 to 45 minutes
No proxy reporting
42,500 respondents: response rate = 75.6%
Languages of interview
English
French
Mandarin
Cantonese
Italian
Punjabi
Portuguese
Vietnamese
Spanish
Content development
Theoretical framework developed
Balance of content, response burden and manageable costs
Operationalization of concepts into workable questions to be asked of a diverse population
Consultation with Advisory Committees
Focus groups, one-on-one testing, pilot test
Basic demographicsFamily / household compositionEthnic self-definitionReligion, languageFamily background and family InteractionSocial networks and civic participationPerceptions of discriminationSense of belonging, trust and satisfactionSocio-economic activities
Final content of the Ethnic Diversity Survey
Survey themes & questionnaire content
Entry
• Age, sex, marital status• Family / household composition
Ethnic self-definition
• Ethnic ancestry• Ethnic identity• Importance of ancestries and identities
Survey themes and content (continued)
Respondent background
• Birthplace
• Citizenship
• Year of immigration
• Other countries lived in
• Visible minority status
• Religion: importance of, participation
Survey themes and content (continued)
Knowledge & use of languages
• First language: understood & spoken
• Knowledge of languages
• Home languages
• Languages used with friends
• Languages used with family to age 15
• Languages used at work
Survey themes and content (continued)
Family background
• Ethnicity, first language, highest level of schooling and religion of mother and father
• Birthplace of parents and grandparents
• Ethno-cultural, immigration, language, education and religion data for spouse
• Language data for child aged 3 or older
Survey themes and content (continued)
Family Interaction
• Frequency of contact with family living in Canada
• Frequency of contact with family living in parents’ & grandparents’ birthplaces and in other countries
• Visits to country of birth & parents’ & grandparents countries of birth
Survey themes and content (continued)
Social Networks
• Friends in ethnic group, up until respondent was age 15 and now
– For 2 highest rated ancestry groups other than “Canadian”
• Importance of carrying on customs and traditions
– For 2 highest rate ancestry groups other than “Canadian” ranked a 4 or 5 in importance in ID module
Survey themes and content (continued)
Civic Participation
• Participation in groups or organizations in the past 12 months (ethnic and other types)
• Frequency of participation (for 3 groups)
• Ethnicity of co-members (for 3 groups)
– For 2 highest rated ancestry groups other than “Canadian” ranked a 4 or 5 in importance in ID module
• Volunteering (for 3 groups)
• Voting in federal, provincial & municipal elections
Survey themes and content (continued)
Interaction with Society
• Feeling uncomfortable because of ethnicity, culture, race, language, religion: up until the age of 15 and now
• Discrimination or unfair treatment in the past 5 years as a result of ethno-cultural characteristics: frequency, reason & place
• Hate crime: experience, reason & worry
Survey themes and content (continued)
Attitudes
• Rating of sense of belonging to family, ethnic group, town/ city/ municipality, province, Canada, North America
Trust & Satisfaction
• General life satisfaction
• Trust: general, family, neighbours, people at work or school
Survey themes and content (continued)
Socio-economic activities
• Highest level of schooling, country of schooling & current school attendance
• Labour force questions, occupation, industry & income (personal & household)
• Ethnicity of co-workers– For 2 highest rated ancestry groups other than
“Canadian” ranked a 4 or 5 in importance in ID module
Ethnic Diversity Survey products
Official release in Statistics Canada’s The Daily September 29, 2003
Analytic article: Ethnic Diversity Survey: portrait of a multicultural society
Public Use Microdata File (PUMF)
Custom tabulations
Analytical file at Research Data Centres
Survey documentation
Survey overview and questionnaire:
http://www.statcan.ca/english/sdds/4508.htm
Codebooks with and without frequencies
Users’ Guide (and methodology and data quality documentation)
WesVar Users’ Guide (bootstrap weights)
Research Data Centres (RDCs)
University of British Columbia University of Calgary University of Alberta University of Manitoba University of Western Ontario University of Waterloo University of Toronto Queen's University Carleton Ottawa Outaouais Local
(COOL) RDC Federal Data Access Centre McMaster University University of Montréal University of New Brunswick Dalhousie University
Analytical data file at RDCs
Access granted through application to Social Sciences and Humanities Research Council, available at:
http://www.sshrc.ca/web/apply/program_descriptions/ciss_reseach_data_e.asp
EDS file includes all content from the survey: raw data and derived variables
Includes detailed geographic identifiers
Includes some 2001 Census information
Final content of the Analytical File
Includes data for 42,476 respondents
All content from the survey (raw data in flat file format with SAS and SPSS cards)
Derived variables
Some 2001 Census information (e.g. major field of study, number of rooms in dwelling)
Postal code and other detailed geographic identifiers are present
Survey weights
Each respondent represents a certain number of other people in the population who were not part of the sample
Population weights & bootstrap weights (used to determine coefficients of variation) are included on the Analytical File present in the RDCS
A weight is associated with each respondent and must be used for all estimates and analysis
The use of survey weights is crucial
The sampling ratio differs widely from one strata to another;
The final weight assigned to each respondent underwent numerous adjustments for non-response and post-stratification;
The weighting of data ensure that the EDS sample is representative of the target population;
Without the weights: false / misleading results for most types of analysis.
WesVar software: variance
Used to verify coefficients of variation (cvs)
Can also be used for regression analysis
WesVar reads .ssd, .xpt, .sav, .dat and .txt files
Does not alter the original file, but creates a new one; can be used to view or print output
Tip: include all the variables in the first WesVar file you create - otherwise you will need to repeat later and importing files can be slow
Analysis: Level of geography
Good quality data are generally available at national, regional and provincial levels and for Toronto, Montréal and Vancouver CMAs;
Atlantic provinces are always aggregated;
Counts at the census subdivision and municipality level are generally small and the results of survey estimations will probably be unreliable and/or the results may be unsuitable for publication because of the risk of statistical disclosure.
Analysis: Complexity of data
Some concepts are similar yet distinctly different from one another: ancestry, identity, visible minority status, language, religion, etc.
Multiple response variables: e.g. ethnic ancestry, identity, languages, organizations, etc.
Follow-up questions for some topics:
(a) universe is restricted;
(b) requires link to inserted ethnic ancestry/language/group or organization.
EDS Public Use Microdata File (PUMF)
Extensive list of variables, but not as detailed as the RDC file… reduced detail protects confidentiality.
Flat file format, with SAS and SPSS cards, CV tables.
PUMF sample was 41,695 compared to 42,476 in the analytic file.
Access through the Data Liberation Initiative, free of charge, or, copies may be purchased for $2,140. To purchase a copy contact client services at (613) 951-5979 or by e-mail at [email protected]
EDS PUMF: Selected content geography and ethnic ancestry
Geography:
• Canada• Toronto, Montréal, Vancouver, Other CMAs,
Non-CMAs
Ethnic ancestry:
• 2 “counter” variables• 2 summary variables• 1 detailed variable (in 8 parts), showing 50
ethnic and cultural groupings
EDS PUMF: Selected content ethnic identity, importance of ethnicity
Ethnic identity:
• 2 “counter” variables• 2 summary variables• 1 detailed variable (in 6 parts), showing 27
ethnic and cultural groupings
Importance ratings for ancestry and identity
Flags indicating differences between original responses versus final codes
EDS PUMF: Selected content ethnic salience
Follow-up questions:
• Raw variables for ethnicity of friends now and when growing up
• Raw and derived variables for ethnic customs and traditions
• Raw variables for knowledge of local ethnic associations
• Derived variables for co-participators have same ethnic ancestry or ethnic ancestries (2 summary variables)
Sense of belonging to ethnic group
EDS PUMF: Selected content birthplace and generations
Respondent’s birthplace: 12 specific countries, 7 broader world regions
Birthplace of parents variable shows “Same as” or “Different from” respondent.
1st, 2nd, 3rd plus generation
Period of arrival for 1st generation: before 1991 or between 1991-2001
Age at arrival for 1st generation: 0-5, 6-14, 15-24, 25-44, 45-64, 65+
Complexity of data on RDC file and PUMF: Concepts & multiple responses
Similar yet distinctly different concepts: ancestry, identity, visible minority status, language, religion, etc.
• Determine relevancy to research, carefully select variables
Multiple response variables: e.g. ethnic ancestry, identity, languages, organizations
• Choose summary or detailed variables
• Detailed variables: need to combine data from more than one variable/field and create user-defined “dummy variables”
Complexity of data on RDC file and PUMF: Follow-up questions
Asked of only a select population;
Requires linking to inserted ethnicity (or language or group or organization);
Ethnicity response provided by the respondent may be slightly different from the text name for the ethnic code included on the file (use of flags recommended);
Again, multiple response issue: need to combine data and create user-defined derived variables.
Research potential with the EDS
Unpacking Ethnicity
EDSEDS
Discrimination & Unfair
Treatment
Social Networks
Participation in Society Transnationalism
Transmission of Culture & Language
Social Capital
Socio-economic Status
Other Statistics Canada data sources on ethnic diversity
Main sources:
Census Longitudinal Immigration Data Base (IMDB)Longitudinal Survey of Immigrants to Canada
(LSIC)
Other sources:Canadian Community Health SurveySurvey of Labour and Income DynamicsYouth in Transition SurveyAdult Literacy and Lifeskills Survey
Questions?
Analytical File:Contact your local RDC analyst
PUMF:Client Services Social & Aboriginal Statistics [email protected]