• Summary
• Downloads
• Explanatory Notes
• Related Information
• Past Releases

EXPLANATORY NOTES


INTRODUCTION

1 This publication presents results from the National Crime and Safety Survey (CSS) conducted throughout Australia during April to July 2005 as part of the Monthly Population Survey (MPS), as a supplement to the April Labour Force Survey (LFS). This is the fourth in the series of Australia-wide self-enumeration Crime and Safety Surveys conducted by the Australian Bureau of Statistics (ABS); previous surveys were conducted in 1993, 1998 and 2002.

2 The publication Labour Force, Australia (cat. no. 6202.0) contains information about survey design, sample redesign, scope, coverage and population benchmarks relevant to the monthly LFS, which also apply to supplementary surveys. It also contains definitions of demographic and labour force characteristics.

3 The CSS collected information from individuals and households about their experience of selected crimes, reporting behaviour and crime-related risk factors. Information was sought from approximately 46,100 persons, of whom 36,500 (79%) responded. Data pertaining to households were sought from approximately 23,200 households, of which 18,600 (80%) responded.

4 Non-response occurs when people cannot or will not cooperate, or cannot be contacted. Non-response can affect the reliability of results and can introduce bias. The magnitude of any bias depends on the rate of non-response and the extent of the difference between the characteristics of those people who responded and those who did not. Examination of responses after follow-up did not indicate any serious problem of non-response bias. The estimation and imputation procedures employed were designed to reduce the effect of non-response (see paragraphs 41-49).

SCOPE OF THE SURVEY

5 The survey was conducted for all persons aged 15 years and over who were usual residents of private dwellings, except:

• members of the permanent defence forces;
• certain diplomatic personnel of overseas governments, customarily excluded for Census and estimated resident population figures;
• overseas residents in Australia;
• members of non-Australian defence forces (and their dependants); and
• visitors to private dwellings.

7 The CSS was conducted in both urban and rural areas in all states and territories, but excluded approximately 120,000 persons living in very remote areas of Australia who would otherwise have been within the scope of the survey. The exclusion of these persons will have only a minor impact on any aggregate estimates that are produced for individual states and territories, except for the Northern Territory where such persons account for around 23% of the population.

COVERAGE

8 The estimates in this publication relate to persons covered by the LFS in April 2005. In the LFS, coverage rules are applied which aim to ensure that each person is associated with only one dwelling, and hence has only one chance of selection in the survey. See Labour Force, Australia (cat. no. 6202.0) for more details.

9 At the end of the April LFS interview, CSS questionnaires were mailed to LFS respondents. Each respondent in the household was asked to complete a questionnaire relating to their personal experience of crime, and return it to the ABS by mail. One questionnaire per household also contained questions relating to experiences of household crime, to be answered by one respondent on behalf of the household as a whole. All questionnaires also included additional questions about sexual assault, to be answered only by persons aged 18 years and over. Completion of the sexual assault questions was voluntary for all persons.

RELIABILITY OF THE ESTIMATES

10 Estimates in this publication are subject to sampling and non-sampling errors:

• Sampling error is the difference between the published estimate and the value that would have been produced if all dwellings had been included in the survey. For further information on sampling error, refer to the Technical Note.
• Non-sampling errors are inaccuracies that occur because of non-response, errors in reporting by respondents and interviewers, and errors made in coding and processing data. These inaccuracies may occur in any enumeration, whether it be a full count such as a census or a sample. Every effort is made to reduce the non-sampling error to a minimum by careful design of questionnaires, intensive training and supervision of interviewers and efficient processing procedures.

12 Country of birth data are classified according to the Standard Australian Classification of Countries (SACC), 1998 (cat. no. 1269.0).

CONTENT

13 Crime victim surveys are best suited to measuring crimes against individuals or households with specific victims, who are aware of and recall what happened to them and how it happened, and who are willing to relate what they know.

14 This survey covered only selected types of household and personal crimes. Household crimes covered in the survey were break-ins, attempted break-ins, and motor vehicle theft. Personal crimes covered in the survey were robbery, assault and sexual assault.

15 Not all types of crime are suitable for measurement by household surveys. No reliable information can be obtained about crimes without specific victims, such as trafficking in narcotics. Crimes of which the victim may not be aware cannot be measured effectively - some instances of fraud and attempted crimes of many types may fall into this category. It may also be difficult to obtain information about some crimes such as sexual offences and assault committed by other household members. Some of these crimes may not be fully reflected in the data collected. No reliable data can be collected by household surveys on crimes against commercial establishments or government agencies.

16 Data on selected crimes recorded by police agencies are available in Recorded Crime -Victims, Australia (cat. no. 4510.0). The CSS provides an additional source of data on crime victimisation for the selected offences, including crime not reported to or detected by the police. This survey identifies the nature of this unreported crime, as well as giving information on victims of repeated crimes, which is not available from police data. The information from the survey should be viewed as a complement to the published police statistics on crime. Reporting rates in the CSS refer to the total number of most recent incidents of an offence that were reported to police expressed as a percentage of the total victims of that offence. Police statistics include victims of all ages, whereas the CSS collects information for persons aged 15 years and over, and sexual assault is only asked directly for persons aged 18 years and over.

17 For this survey, the definition of total victims is restricted to those offences included in the survey; it by no means represents total crime.

Statistical measures of crime victimisation

18 The level of victimisation can be measured in more than one way. The most common measure derived from crime victim surveys is prevalence, that is, the number of the relevant population that have been a victim of a given offence at least once in the reference period. Victimisation prevalence rates are used in this publication, and are generally expressed as a proportion of the total relevant population.

19 Another measure that can be used is incidence. This is the total number of incidents of the offence that occurred in the reference period. As some victims experience repeated incidents of victimisation, incidence numbers are typically higher than prevalence numbers. Incidence rates are also used in this publication, and are also generally expressed as a percentage of the total relevant population.

DATA QUALITY

Interpretation of results

20 Information recorded in this survey is essentially ‘as reported’ by respondents and hence may differ from that which might be obtained from other sources or via other methodologies. This factor should be considered when interpreting the estimates in this publication.

21 Information was also collected on respondents’ perceptions of problems in their neighbourhood. Perceptions are influenced by a number of factors and can change quickly. Care should therefore be taken when analysing or interpreting this data.

Significance testing

22 In tables 1, 2, 6 and 15 of this publication, apparent changes in results between the 2005, 2002 and 1998 surveys have been tested to determine whether the changes are statistically significant. That is, to determine whether it is likely that the differences observed in sample estimates indicate real differences in the population. In these tables, cells which have not changed significantly over time are indicated. See Technical Note (paragraphs 13-14) for further details.

Sexual assault data

23 Response rates for sexual assault in 2005 are lower than in previous years. This is most likely due to changes made to the survey methodology, which included combining the sexual assault questions with the main survey form, and some changes to the structure and wording of the screening questions. Due to the low response rates for sexual assault only limited data is available for 2005. For further information on sexual assault non-response and imputation, see paragraph 44.

Reasons for feeling unsafe

24 For those respondents who indicated in the survey that they felt unsafe or very unsafe, either during the day or after dark, data was also collected on their reasons for feeling unsafe. This data has not been collected in previous ABS crime and safety surveys and due to limited testing prior to the survey, the estimates produced are regarded as 'experimental'. The data therefore is not included in this publication but is available on request.

DATA COMPARABILITY

Comparability of time series

25 Revisions are made to population benchmarks for the LFS after each five-yearly Census of Population and Housing. The last such revision was made in February 2004 to take account of the results of the 2001 Census of Population and Housing. Estimates from supplementary surveys conducted from February 2004 are based on these population benchmarks. However, as the 2005 CSS includes both household and person level estimates, these estimates are based on 1996 Population Census benchmarks.

Comparability with previous ABS Crime and Safety Surveys

26 Since 1990, the ABS has been conducting state-specific Crime and Safety Surveys in New South Wales (annually), Western Australia (1991, 1995, 1999 and 2000), South Australia (1991, 1995 and 2000), Victoria (1994 and 1995), Tasmania (1998) and Queensland and the Australian Capital Territory (1995), often with funding from the state concerned. A national survey was last conducted in 2002, and previous to that in 1998, 1993, 1983 and 1975. All of the surveys since 1990 have required respondents to complete the questionnaires themselves and mail these back to the ABS.

27 While an important use of crime victims surveys is to establish the trend of crime and reporting behaviour over time, care must be exercised in the comparison of the results of the 1993 survey with the 1998, 2002 and 2005 surveys for robbery and assault because of significant definitional differences. The household crime definitions are comparable between the surveys, however, changes in respondents’ attitudes towards crime in the last twelve years may have affected their responses to the surveys.

Comparison with other ABS surveys

28 There are different crime statistics collections and these can yield differing results. Caution should be taken in comparison across other ABS surveys that address crime and justice issues. For more information on comparisons with other surveys, please refer to Information Paper: Measuring Crime Victimisation, Australia - The Impact of Different Collection Methodologies, 2002 (cat. no. 4522.0.55.001).

29 A review of Australian crime surveys has commenced. The context for the review is outlined in the National Information Development Plan for Crime and Justice Statistics (ABS cat. no. 4520.0).

30 The broad objectives of this review are to:

• where possible, consolidate national needs for crime survey data in Australia;
• identify deficiencies and overlaps in current data collections with reference to national data requirements; and
• develop strategies to improve integration and coordination in national crime survey data.

32 Findings from the second stage of consultation will guide planning for crime surveys in the ABS forward work program.

Comparison with police statistics

33 The terms used for the offences (such as robbery, assault) may not necessarily correspond with the legal or police definitions which are used for each offence. This is because responses obtained in this survey are based on the respondent’s perception of their having been the victim of an offence. The definitions of these terms which are used in this survey are based on the wording of the questions asked of the respondent, and can be found in the Glossary of this publication.

34 For each offence type, survey respondents were asked how many of the incidents that happened to them in the previous 12 months became known to the police. However, there is no verification that a crime report has actually been made for the offence. There are many reasons why an incident which is reported to police may not be recorded as a crime. Victims may advise police of a matter but not seek to have it dealt with as a crime. This is especially true of incidents where the offender is related, or otherwise known, to the victim. The information received by police may not be sufficient to allow them to determine whether a crime has been committed, or what kind of offence has been committed. The incident may also have been coded to a different offence category.

35 Another source of variation between the survey results and crimes recorded by police relates to differences in scope. This survey collects information on the personal crimes of robbery and assault for all persons aged 15 years and over, and sexual assault for persons aged 18 years and over; whereas police statistics include victims of all ages and comparisons need to allow for this. Police record all crimes reported to them in the reference period, whether or not the incidents took place during that period or at an earlier time, and count all distinct estimates of victimisation reported.

36 It is also possible that respondents to the survey may have included some incidents which occurred outside the 12 month period. Information was collected on which quarter in the last 12 months the most recent incident occurred, in an attempt to reduce this effect.

Other methodological issues

37 In the interpretation of the results of the 2005 survey, consideration should be given to the representativeness of the sample. This is affected by the response rate and also the fact that the survey only covers people living in private dwellings. It does not include such people as the homeless or people living in special accommodation; sections of the community which may be heavily victimised. Also, how respondents complete their questionnaire will depend on their personal circumstances, their interpretation of the questions, and how much they are willing to divulge.

38 Self-enumeration questionnaires may affect the willingness of a person to reveal personal information on issues such as assault and sexual assault. There may also be problems if the respondent’s literacy skills are not adequate to understand a reasonably complex form.

39 In order to derive the total number of incidents for the selected offences the survey looked at victimisation as discrete incidents, whereas for many victims of violence it is an ongoing situation. The result of such treatment is an undercount of total incidents in this survey.

DATA PROCESSING

40 Consistency and range edit checks along with careful clerical scrutiny were applied to all forms, mainly during data entry, to ensure that answers were complete and consistent within the household or person record. The survey included questions asking respondents to describe the incident, and these descriptions were used to check that other questions had been consistently answered. Where an incident had been incorrectly included as a particular offence type, then the incident was either recoded or discounted, as appropriate. This either resulted in an adjustment to the number of incidents, or the respondent being counted as a non-victim of the original offence.

Non-response and imputation

41 There were a number of cases where the questionnaires were not complete. The most common problem was when questions that should have been answered had been left blank.

42 Missing data were treated in one of the following ways:

• where possible, missing information was imputed from other answers on the form, following a standard set of imputation rules as derived for this survey
• where the data could not be imputed, a ‘not stated’ code was used
• forms with significant amounts of missing data were treated as non-response.

44 Completion of the sexual assault questions was voluntary. Of those respondents aged 18 years and over who completed the main survey form, 75.4% of females and 71.5% of males completed the sexual assault questions. The number of sexual assault victims for non-respondents was imputed for Table 6. The imputed number of victims was based on the assumption that the victimisation rates were equal for respondents and non-respondents within age group and sex categories. Apart from the imputed number of victims, no further imputation was carried out for subsequent sexual assault questions, and any tables providing more detail on sexual assault have these questions coded as 'not stated'.

Non-response bias adjustment and weighting of data

45 A complex set of multistage weighting procedures was employed to adjust for non-response and improve the precision of estimates in this survey.

46 Estimates derived from this survey were obtained in two stages. In the first stage, adjustment for non-response was carried out based on the demographic composition of the MPS sample, as well as on results of the analysis of the effect of reminder action on the responses obtained.

47 In most surveys there is little information available about non-respondents. In this survey, however, demographic characteristics of most non-respondents were available from the MPS which had a very high response rate. Weighting adjustments were carried out using the distribution of the demographic characteristics from the MPS which were correlated with crime victimisation.

48 For this survey there were two separate reminder follow-ups (by mail for the first reminder, and mail or telephone for the second reminder). Indications of bias due to the difference between respondents and non-respondents were investigated using successive waves of responses. Any significant bias identified was adjusted using data from later waves of responses.

49 The second stage of the weighting procedure was a complex ratio estimation procedure which ensured that the survey estimates conformed to an independently estimated distribution of population by age, sex, and part of state, rather than to the age, sex and part of state distribution within the sample itself. The procedure also ensured that household estimates conformed to the independently estimated distribution of households by certain household characteristics (number of adults and children in the household) rather than to the distribution that existed among the responding households.

PRODUCTS AND SERVICES

50 Summary results from this survey, compiled separately for each state and territory, will be available in spreadsheet form from the ABS website <www.abs.gov.au> or on request to the ABS.

51 For users who wish to undertake more detailed analysis of the survey data, an expanded confidentialised unit record data file (CURF) is proposed to be available in mid 2006. The Expanded CURF will be accessible through the ABS Remote Access Data Laboratory (RADL) system. Further information about these files, including details of how they can be obtained and conditions of use, will be available on the ABS website <www.abs.gov.au>.

52 Special tabulations are available on request. Subject to confidentiality and sampling variability constraints, tabulations can be produced from the survey incorporating data items, populations and geographic areas selected to meet individual requirements. These can be provided in printed or electronic form. A list of data items available from the survey is available in Appendix 2. Further information about the survey and associated products can be obtained from the contact officer listed at the front of this publication.

ACKNOWLEDGEMENT

53 ABS surveys draw extensively on information provided by individuals, businesses, governments and other organisations. Their continued cooperation is very much appreciated: without it, the wide range of statistics published by the ABS would not be available. Information received by the ABS is treated in strict confidence as required by the Census and Statistics Act 1905.

NEXT SURVEY

54 Subject to the ABS's review of Australian crime surveys, the ABS plans to conduct this survey again in 2008.

RELATED PUBLICATIONS

55 Other ABS publications which may be of interest are shown below. Most of these are available at <www.abs.gov.au>

• Australian Standard Offence Classification, 1997 (cat. no. 1234.0)
• Community Safety, Tasmania, October 1998 (cat. no. 4515.6)
• Crime and Safety, Australia: Supplementary National and Standard Tables, 1998 (cat. no. 4509.0.40.001)
• Crime and Safety, Australia, April 2002, 1998 and 1993 (cat. no. 4509.0)
• Crime and Safety, Australia: Expanded Confidentialised Unit Record File, 2002 (cat. no. 4509.0.55.002)
• Crime and Safety, New South Wales, April 2004 (cat. no. 4509.1)
• Crime and Safety, New South Wales and the Australian Capital Territory, April 1995 (cat. no. 4509.1)
• Crime and Safety, Queensland, April 1995 (cat. no. 4509.3)
• Crime and Safety, South Australia, October 2000 (cat. no. 4509.4)
• Crime and Safety, Victoria, April 1995 (cat. no. 4509.2)
• Crime and Safety, Western Australia, October 1999 (cat. no. 4509.5)
• General Social Survey, Australia, 2001 (cat. no. 4159.0)
• Information Paper: Measuring Crime Victimisation, Australia - The Impact of Different Collection Methodologies, 2002 (cat. no. 4522.0.55.001)
• National Aboriginal and Torres Strait Islander Survey: Law and Justice Issues, 1994 (cat. no. 4189.0)
• Personal Safety Survey, Australia, 2005 (cat. no. 4906.0) (due for release July 2006)
• Recorded Crime - Victims, Australia, 2004 (cat. no. 4510.0)
• Victims of Crime, Australia, 1983 (cat. no. 4506.0)
• Western Australian Statistical Indicators, March 2001 (cat. no. 1367.5)
• Women’s Safety, Australia, 1996 (cat. no. 4128.0)

NON-ABS PUBLICATIONS

57 Non-ABS sources which may be of interest can be accessed through the National Centre for Crime and Justice Statistics theme page on the ABS web site <www.abs.gov.au>.