Rules of engagement

South African Social Attitudes Survey (SASAS) Rules of Engagement

  1. The South African Social Attitudes Survey (SASAS) will be cross-sectional in design, though there is scope for specialised follow-up ad hoc or panel studies outside the scope of SASAS and using sub-samples.
  2. SASAS will be conducted on an annual basis.
  3. The envisaged sample size is 5000 individuals, subject to Prof. Stoker's input and budget considerations.
  4. There will be two versions of the survey instrument in any given year. Some modules will be included in both instruments, but other modules will be included in only one of the instruments.
  5. SASAS has been designed to incorporate clusters of questions in the form of modules rather than individual questions. This is intended to discourage ad hoc questions that are better suited to Omnibus-type surveys.
  6. All questions must be designed with the intention of being repeated.
  7. SASAS will focus on underlying (climate) changes rather than topical (weather) changes.
  8. Not every core question or core module needs to be asked every year.
  9. All modules will be considered for compression in future years.
  10. Continuity in the chosen measures over time is a fundamental priority. Wherever salient and sensible, an attempt will be made to extend the time series backwards. Nonetheless, this does need to be considered against the appropriateness of the question at the time of implementation. Should a question be judged to be wholly unsuitable, then a ruling in favour of a change to the chosen measure should be adopted.
  11. Each module, other than the core demographic module and the ISSP module, should not exceed 10 minutes in duration.
  12. Each survey instrument should aim to be on average 60 minutes in duration.
  13. In addition to the core demographic module (15 minutes), there will be a maximum of 10 different subject areas/modules within any version of the questionnaire. At present, we are estimating four questions per minute.
  14. Within the scope of the time limits outlined above, there is a need to ensure that each module is developed so as to attain sufficient depth in relation to the subject matter being addressed.
  15. Modules that are intellectually inter-related need to be included in the same version of the questionnaire in any given year.
  16. Each annual questionnaire should contain a range of policy areas and disciplinary boundaries.
  17. The International Social Survey Programme (ISSP) module, which typically includes approximately 60 questions and takes 15 minutes to conduct, will be included as an annual module. It is intended that the ISSP module will be sustained using a split sample.
  18. All questions should be understandable to all respondents. It is therefore paramount that jargon be excluded from the questions.
  19. No open questions are to be included in the survey instruments, with the possible exception of the demographic module.
  20. SASAS does not contain client modules in the sense of clusters of questions specified by outside users. If externally funded modules are included, the researcher rather than the funder has the final say. Other periodic instruments, whether cross-sectional or panel in design, will be developed for subscriber use.
  21. The SASAS Project Team will run the survey series. However, the head of the SASAS Project Team has the final say on the contents.
  22. The SASAS Project Team will be responsible for consulting, commissioning expert papers and deciding on the final content of the product.
  23. Each member of the SASAS Project Team needs to be present for the questionnaire design workshop, that is scheduled for February each year. A full-time Project Manager will be appointed to SASAS.
  24. For the first year of surveying, internal fieldwork capacity will be developed such that outsourcing will not be required. This will consist of full-time regional coordinators and part-time field teams (quality assurers/supervisors and fieldworkers).
  25. SASAS Project Team members will undertake supervision of the regional coordinators and conduct random backchecks.
  26. Members of the Project Team would be held to rigorous pretesting, interviewer debriefings and piloting. Moreover, the SASAS Project Manager would be responsible for personally conducting several interviews in the field and undertaking site visits to the various sampling points.
  27. The SASAS questionnaires should aim to be administered rather than self-completed.
  28. Subject specialists may be invited to help devise modules and write chapters of the annual SASAS publication, wherever possible with SASAS Project Team members as co-authors.
  29. HSRC researchers will get access to data from all the modules included in the survey prior to official release.
  30. Final editorial control of the annual SASAS publication rests with the editor rather than with the specialists.
  31. A reference and editorial board committee will be established to decide on, amongst other things, the pre-release and release of results in various media.
  32. Complicated terminology should be avoided in the publication and sophisticated statistical analysis should be relegated to appendices.
  33. The release of the data in the public domain must coincide with the release of the annual SASAS publication.
  34. A Strategic Framework for SASAS needs to be developed by the Project Team, which will plan for, inter alia, the thematic content and sequencing for the first seven to ten years of the survey series. Moreover, the Project Team will commit to updating the Framework on an annual basis.