Online credit card fraud against small businesses

Kate Charlton and Natalie Taylor
ISBN 0 642 53846 8 ; ISSN 1326-6004
Canberra: Australian Institute of Criminology: 2004
(Research and public policy series, no. 60)

Section 2 : Overview of AIC online credit card fraud survey

There is no centralised register of businesses that currently trade online from which to draw a sample for survey purposes. However it was assumed that if people purchase a particular type of good over the internet more often than other types of goods, then retailers who sell those products may be more likely to trade online. Therefore, five business types that have been identified as more popular with internet shoppers were included (Adey 1999; Shimmin 2000; US Census Bureau 1999). Businesses within each of these types were randomly selected from Australia On Disc (the electronic version of the Yellow Pages directories). The businesses were stratified by the five selected business types (see Appendix 1 for detailed information about the methodology). Only businesses which accepted credit card payments on their own web site were classified in this study as trading online.

The questionnaire was administered over the telephone by a Roy Morgan trained interviewer, using the computer-assisted telephone interviewing technique (CATI). The questionnaire is provided in Appendix 2. At least three call-backs were made to those businesses which could not be contacted, with more if necessary to achieve the required samples. The average length of time to complete the questionnaire was 11 minutes.

In total 1,078 small businesses were interviewed, evenly distributed between the five business types. The overall response rate was 62 per cent. For purposes of better representing the retailers in these five types across Australia, the data were then weighted by business type (see Appendix 1), providing a total number of small businesses in these five business types of 6,657 across Australia.

In this report, weighted data have been used when providing descriptive statistics to indicate estimated frequencies for all retailers in the five business types across Australia (for example, losses, number of incidents and so on). Where statistical analyses are conducted, the effective sample size (weighted data for original sample) has been used to ensure statistical significance has not been inflated. Therefore, in this report:

• weighted data are used to provide descriptive information for the population and this is indicated when numbers are given for retailers 'across Australia'; and
• where tests of significance are calculated (for example, regression analyses), the sample has been weighted for better representativeness, but scaled down to the effective sample size.

Demographics of businesses included in this survey (including sex of the retailer, business size, annual turnover and remoteness of business location) are provided in Appendix 1.