Home → Publications → Reports → Technical and background paper series → 16 → Analysis of the sample (in: The Australian component of the 2004 International Crime Victimisation Survey)

The Australian component of the 2004 International Crime Victimisation Survey

Graham Challice and Holly Johnson
ISBN 0 642 53889 1 ; ISSN 1445-7261
Australian Institute of Criminology, 2005
(Technical and background paper series, no. 16)

Download

PDF print version: analysis of the sample (PDF 164kB)
PDF print version: full report (PDF 560kB)

Analysis of the sample

The importance of random selection of respondents for a survey on crime victimisation is highlighted in the following analysis. First, respondents with telephone numbers that could be matched to the electronic White Pages were compared with numbers that could not be matched, then respondents reached within eight call attempts were compared with those reached at nine attempts or more. The results highlight the importance of random selection and an extended call regime.

4.1 Matched and unmatched numbers

Table 5 outlines differences between respondents with telephone numbers that could be matched back to the electronic White Pages (and were subsequently sent an approach letter) and respondents with unmatched numbers (which includes silent and newly listed numbers). Profiles were compared by selected demographic characteristics and questionnaire items to examine the impact of choosing an RDD sample frame for the ICVS. Respondents from unmatched listings are significantly different from those with a matched status for all selected demographic variables, with the exception of gender. Of particular note is the proportion of interviews achieved from the unmatched sample with persons under 34 years of age, single, divorced/separated persons and persons newly arrived at their current postcode. This is important for the ICVS, since factors such as age, marital status, speaking a language other than English at home and dwelling type have been shown to be related to crime victimisation measures and attitudes to crime.

Table 5 : Analysis of demographic characteristics, by listing status
Match type	Matched	Unmatched
Number	4730	1270
Demographic characteristics
Gender
Male	46.7	45.8
Female	53.3	54.2
Age
16-24	14.8	17.4^*
25-34	16.7	24.0^**
35-54	35.6	39.4^*
55-64	15.3	10.5^**
65 plus	17.6	8.7^**
Speaks language other than English at home	13.6	16.6^**
Marital status
Single	27.1	32.3^**
Married	50.1	42.3^**
Living together as a couple (but not married)	6.4	9.0^**
Divorced/separated	10.4	13.4^**
Widowed	6.0	3.0^**
Main activity
Doing paid work	55.1	58.2^*
Looking for work	1.8	4.7^**
Doing home duties	10.2	13.2^**
Retired/on a pension	22.7	14.4^**
Studying	9.6	9.1
Other	0.5	0.6
Dwelling
Flat or apartment	12.6	16.8^**
A terraced house	3.9	4.1
A freestanding house	83.3	79.0^**
Other	0.2	0.1
Length of time lived at postcode
Less than a year	8.3	17.3^**
1 or more years, up to 3 years	18.8	20.4
3 or more years, up to 5 years	11.7	13.6
5 or more years, up to 10 years	16.7	18.0
10 years or more	44.5	30.7^**
^* statistically significant p<0.05 ^** statistically significant p<0.01 Source: Australian Institute of Criminology, International Crime Victimisation Survey, 2004 [computer file]

There were also some significant differences in crime victimisation and attitudinal questions according to matched status (Table 6). Respondents from unmatched households were significantly more likely to have been a victim of all but one types of crime. They were also significantly less likely to feel safe walking alone in their area after dark, or waiting for or using public transport after dark, and more likely to have seen evidence of drug use in their area.

Table 6 : Analysis of questionnaire items, by listing status
Match type	Matched	Unmatched
Base	4730	1270
Crime victimisation	%	%
Motor vehicle theft	6.6	8.7^*
Theft from motor vehicle	20.6	22.7
Motorcycle theft	5.3	4.0^*
Bicycle theft	9.8	12.3^*
Burglary	12.2	15.5^**
Attempted burglary	9.2	13.0^**
Robbery	3.4	5.7^**
Personal theft	14.0	16.7^*
Assault/threat	17.5	24.4^**
Victim of at least one crime	51.3	58.0^**
Attitudes to crime	%	%
Feels safe walking alone in area after dark	72.5	69.0^*
Feels safe waiting for/using public transport after dark	60.7	55.9^**
Likely/very likely that house will be broken into within the next 12 months	38.1	38.4
Very/somewhat worried about being threatened/assaulted due to ethnicity	9.5	11.7^*
Often/from time to time - evidence of drug use in area in the last 12 months	26.7	35.0^**
^* statistically significant p<0.05 ^** statistically significant p<0.01 Source: Australian Institute of Criminology, International Crime Victimisation Survey, 2004 [computer file]

4.2 Analysis by call attempt

An extended call regime was used to contact households in the ICVS. As a result, a substantial proportion (one in five respondents) were interviewed on the ninth or more call attempt. An analysis of response was undertaken by the call number on which the interview was achieved to examine whether respondents interviewed late in the ICVS call cycle were different in any way from those interviewed within a standard call cycle. As shown in Table 7, the extended call cycle contributes to the improved representation of persons who are young, single, working, speak a language other than English at home and live in an apartment or flat. Persons interviewed late in the call cycle were also significantly more likely to have been a victim of crime and to reside in an unmatched household.

Table 7 : Analysis of response, by call attempt
Call attempt on which interview was achieved	1 to 8	9+
Base	4827	1173
Demographic characteristics
Age
16-24	14.3	19.4^**
25-34	17.1	22.8^**
35-54	36.3	37.8
55-64	14.7	12.3^**
65 plus	17.6	7.7^**
Speaks language other than English at home	13.3	18.3^**
Marital status - single	26.5	35.5^**
Main activity - doing paid work	53.1	66.8^**
Dwelling type - flat or apartment	12.6	17.2^**
Length of time lived at postcode - less than three years	28.2	34.1^**
Victim of at least one crime	52.1	55.5^*
Listing status
Unmatched sample	20.3	24.9^**
^* statistically significant p<0.05 ^** statistically significant p<0.01 Source: Australian Institute of Criminology, International Crime Victimisation Survey, 2004 [computer file]

Next section: Survey weights
Previous section: Sample design and selection
Contents