A total of 1,507 CATI (Computer Assisted Telephone Interviewing) interviews were conducted nation-wide in May 2004. In addition, the results obtained for three questions were verified by re-asking these questions to 1214 respondents as a part of Roy Morgan Research's CATIBUS, a multi-client, shared cost survey. The telephone numbers for both surveys were randomly selected from the latest version of the Electronic White Pages with quotas set according to age, sex and location. Interviews were conducted from Roy Morgan Research offices in Melbourne and Adelaide, with interviewer briefings conducted by Field Managers at each of the sites.
This report deals with the stand-alone telephone survey, referred to as the 2004 Privacy Survey. The CATIBUS Verification project is reported in Appendix Two.
The questionnaire was designed in close consultation with staff from the Office who, in turn, sought input from a panel of stakeholders. Questionnaire design was aided by the findings from the 2001 study, with the majority of questions repeated from the 2001 study. Whilst many questions were repeats of the 2001 study, their order in the 2004 questionnaire differed. This reordering was to minimise the impact one question may have on the results obtained for subsequent questions. A copy of the 2004 questionnaire is attached at Appendix One.
After finalisation of the draft questionnaire, a pilot phase of ten interviews was conducted to ensure the questionnaire worked well over the phone. Following the pilot, some minor changes were made to the questionnaire, to improve questionnaire flow and respondent comprehension.
The average 2004 interview length was 23 minutes.
To ensure the sample adequately represented a true cross-section of the Australian population it was stratified by location (capital cities, other metropolitan and rural) and state.
Age and sex quotas were applied within capital city and non capital city quotas for each state to ensure proportional representation.
The number of interviews conducted within each location is detailed in the following table.
| Age/Sex | Total | Syd | Rest NSW / ACT | Mel | Rest VIC | Bris | Rest QLD | Adel | Rest SA / NT | Perth | Rest WA | TAS |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Male 18-24 | 91 | 16 | 13 | 20 | 7 | 10 | 6 | 6 | 1 | 8 | 2 | 2 |
| Male 25-34 | 136 | 30 | 17 | 26 | 8 | 13 | 15 | 7 | 4 | 10 | 3 | 3 |
| Male 35-49 | 223 | 49 | 32 | 40 | 14 | 19 | 22 | 13 | 6 | 17 | 6 | 5 |
| Male 50+ | 283 | 58 | 41 | 48 | 21 | 23 | 31 | 17 | 8 | 22 | 7 | 7 |
| Female 18-24 | 93 | 24 | 9 | 21 | 6 | 8 | 8 | 5 | 1 | 7 | 2 | 2 |
| Female 25-34 | 153 | 38 | 18 | 31 | 8 | 13 | 14 | 8 | 3 | 12 | 4 | 4 |
| Female 35-49 | 222 | 46 | 29 | 40 | 14 | 20 | 25 | 12 | 8 | 17 | 6 | 5 |
| Female 50+ | 306 | 63 | 45 | 55 | 22 | 26 | 32 | 19 | 8 | 21 | 7 | 8 |
| Total | 1507 | 324 | 204 | 281 | 100 | 132 | 153 | 87 | 39 | 114 | 37 | 36 |
Base - All Respondents n=1,507
A total of 200 interviews (13%) were audited by CATI supervisors as part of Roy Morgan Research's quality control procedures. While two-way 'live' auditing (using a CATI system that allows the supervisor to listen in and watch the interview process as it is happening) was undertaken in the Melbourne office, one-way audits were conducted by the Field Manager in the Adelaide Office. (One-way auditing is similar to two-way auditing however the respondent cannot be heard.)
No difficulties or problems with interviewers or the interviewing process were revealed through the auditing process.
The following table shows the number of calls made to achieve the 1,507 interviews, along with the number of refusals and terminations.
| Response | Number |
|---|---|
| Interviews achieved | 1,507 |
| Refusals | 8,642 |
| Terminated mid-interview (respondent drop out) or terminated due to communication difficulty | 1,007 |
| Quota fail (i.e. no-one in household meets criteria. This would occur towards the end of the survey when many of the age quotas had already been filled) | 1,599 |
| Number called 4 times and no answer or engaged on each occasion | 2,230 |
| Appointments, engaged, no answer (called 1-3 times) | 2,275 |
| Unobtainable (number invalid, fax, data or no longer in use) | 4,475 |
Base - All Respondents n=1,507
Of all households where contact was made with an individual, approximately one in seven completed the interview, hence a response rate of 14%. Given the length of interview, this response rate is acceptable and similar to response rates of other comparable surveys.
The call results which were included in calculating the response rates were:
There is an increasing trend for more households to refuse to respond to surveys. Invasion of privacy and being too busy are the main reasons given for these refusals. Whilst some surveys require precise, hard measures (and a high response rate), this survey is seeking comparative levels of concern and relationships between highly context dependent 'soft' measures. The response rate achieved for this survey is within acceptable industry boundaries for this type of survey.
Attempts to increase the response rate by more intensive follow ups or pre-survey contact would have required a substantially larger injection of public money. It is deemed unnecessary to ascertain measurements of this precision for soft measures.
The data collected in the survey were weighted according to age, location (including state) and gender characteristics of the wider Australian population with estimates derived from the latest ABS figures.
Unless specifically stated, percentages quoted throughout the report are weighted and refer to estimates relating to the Australian population aged 18 years or over. It should also be noted that percentages have been rounded to the nearest full number and may not, in some instances, add to exactly 100 percent.
The following report provides a descriptive analysis of each survey question. Results are analysed across a number of relevant variables including age, location, state, education level and income level.
In general, the sampling variation for the sample at the national level for the 2004 Privacy Survey is 1.1% to 2.5%, i.e., there is a 95% probability that the same results within a +/-2.5% range would be obtained.
^top : Last Updated : 27 October 2004