National Statistics Omnibus Survey 1990-2000 by wuxiangyu

VIEWS: 14 PAGES: 23

									 National Statistics
 Omnibus Survey


    Nina Parry-Langdon
Office for National Statistics
        11 May 2007              
                                 
                                 
                                 
   ONS Social and Vital Statistics
         Division (SVS)

• SVS relocated to ONS HQ in Newport
• The NS Omnibus is one of the social
  surveys in SVS (e.g. GHL, LFS, FES)
• Experienced in running large scale random
  sample household surveys

                                    
                                    
                                    
                                    
             Omnibus Survey
•   Background
•   Who uses the NS Omnibus
•   Why and what is it used for
•   Maintaining quality
•   Survey design
•   How can it be used?

                                  
                                  
                                  
                                  
                Background

- Running since 1990
- 12 survey months since April 1995
- Random sample survey
- Good response rate
- Fast data delivery
- Maintenance of data quality

                                      
                                      
                                      
                                      
 Who can use the NS Omnibus?

• Non-profit making organisations
  - government departments
  - academics
  - charities


                                    
                                    
                                    
                                    
  Why use the NS Omnibus?
The Omnibus Survey is a multipurpose
survey that offers a fast, cost-effective and
reliable way of obtaining information on a
variety of topics too brief to warrant surveys
of their own



                                      
                                      
                                      
                                      
 What is the Omnibus used for?
- Measuring public awareness of new policies
- Providing quick answers to questions of
  immediate policy interest
- Measuring the effect of publicity campaigns
- Question testing and piloting
- Providing a sample of respondents for follow-up
  investigations (subject to respondents' consent)
     note: personal information is not linked to
           survey responses
                                      
                                      
                                      
                                      
                        Topics
•   Public confidence in official statistics
•   Contraceptive use and sexual health
•   Smoking and drinking behaviour
•   Cooking behaviour
•   Food safety
•   Fire safety
•   Gambling
•   E-society
•   Pensions                                   
                                               
                                               
                                               
              Omnibus services
• A focus on customer care including advice on
  question design if needed

• Client outputs
    - technical report
    - frequency counts
    - tables
    - over 60 classificatory variables
    - the complete dataset
    - sampling errors                      
                                           
                                           
                                           
     Additional Omnibus services

•   Derived variables
•   More complex data analysis
•   Interpretative report
•   Publication



                                 
                                 
                                 
                                 
     Maintaining Data quality
Three main areas:

1. Sample/weighting
2. Data collectors
3. Data production

                            
                            
                            
                            
                    Sample
• Random probability sample
• Sampling frame - PAF ‘small users’
• Sampled area - Great Britain
• Sample size - each month 67 postal sectors are
  selected
         - 30 addresses at each sampling point
         - 1 adult per household at random at
          each address
• Achieved - around 1800 addresses
  each survey month                   
                                      
                                      
                                      
               Weighting
From April 2005:

•   New weighting system to correct for non-response
    bias
•   Gross up data by age/sex/region
•   Population control totals used on the LFS
•   Improve precision for most variables
•   Weights supplied at a person level for each month
•   Sample errors provided

                                              
                                              
                                              
                                              
            Data collectors
• From general SSD pool
  - rigorous recruitment
  - 4-day interviewer training programme
• Specific Omnibus training
• Post-briefing exercise
• Support from field based regional managers
• Support from HQ field staff
• Inexperienced interviewers restricted

                                    
                                    
                                    
                                    
           Data production
• Computer Assisted Personal Interviewing
  (CAPI)
• Data checking
• Data cleaning
• Recoding
• Table checking

                                    
                                    
                                    
                                    
             Quality Indicators

•   Omnibus monthly response rate
•   Individual interviewer’s response rates
•   Interviewer experience
•   Length of interview
•   Number of reissues
•   Amount of post fieldwork editing
•   Timeliness of outputs
                                          
                                          
                                          
                                          
                     Cost
• Different costs for:
  – simple pre-coded questions
  – showcards
  – multi-codes


• Discounts for:
  – repeated modules
  – questions asked of e.g. < 30% of the sample


                                           
                                           
                                           
                                           
            Survey Timetable
Example survey month:                         June
• Closing date for final questions          14 May
• End of programming                        29 May
• Field testing of questionnaire    29 May - 6 June
• Questionnaire scatter                      7 June
• Survey fieldwork dates           11 June - 7 July
• Data ready for Research          17 July - 19 Aug
• Data & reports delivered to clients       20 Aug

                                          
                                          
                                          
                                          
How can the Omnibus data be used?
Omnibus clients:


•   Trend analysis (repeat modules)

•   ‘snapshot’ of opinions/awareness

•   Merge datasets for in-depth analysis


                                       
                                       
                                       
                                       
How can the Omnibus data be used?
Recent Omnibus reports:

• DoH/Information Centre for health and social
  care:
  - Contraception and Sexual Health, 2005/06
  - Drinking: adults behaviour and knowledge in 2006
  - Smoking-related behaviour ad attitudes, 2005

• Department for International Development:
  - Public attitudes to development 2006

                                                 
                                                 
                                                 
                                                 
                         Dissemination
Press releases
e.g. Report on Contraception and Sexual Health, 2005/06

   Headline findings include
• One fifth (20 per cent) of women aged 16-49 had undergone a test for
  Chlamydia at some point in the past, of whom 27 percent had done so in
  the previous year

• Awareness of Chlamydia is increasing. The proportion of people who
  recognised that it is a sexually transmitted infection (STI) has increased
  sharply since the question was first asked in the 2000/01 survey from 35
  per cent to 79 per cent of men and from 65 per cent to 91 per cent for

                                                                  
                                                                  
                                                                  
                                                                  
  women.
         Omnibus archived data

•   One year after the survey month
•   Public access via the Data Archive
•   Non-disclosive data
•   Restricted data


                                         
                                         
                                         
                                         
               Contacts

• Omnibus telephone query: 01633 655810
• E-mail: Omnibus@ons.gov.uk
• Website:www.statistics.gov.uk




                                  
                                  
                                  
                                  

								
To top