Road safety issues for people from non-English speaking backgrounds

Monash University Accident Research Centre - Report #176 - 2000

Full report in .pdf format [313KB]

Authors: N. Haworth, M. Symmons and N. Kowadlo

Abstract:

Research investigating the differences between road users from a non-English speaking background (NESB) and the English speaking population (ESB) is inconclusive. Other variables such as socio-economic status, education level, employment, residence in areas of high population density and duration of residence may be as important or more so as a predictor of crash involvement than ethnicity. Statistics should also be analysed to take account of exposure, such as unit of travel or number of trips taken.

As crash statistics databases maintained by the Police do not include an ethnicity variable, the aim of the project was to examine crash involvement of NESB groups using other datasets. Hospital injury databases code preferred language and country of birth, but contain too many 'unknowns'. Other Australian Hospital datasets were investigated, but were not be coded any more completely than Victoria's or the number of NESB cases was too small. The National Coronial Information System, various comprehensive and third-party insurance databases, and ABS data sources were also not able to provide useful or complete information to determine the crash involvement of NESB groups.

Generating a more complete set of hospital data is the most attractive option for collecting data, possibly concentrating on a particular hospital or network in a high density NESB location. The time needed to accumulate sufficient data would be dependent on the number of hospitals participating and the percentage of their clients who can be classified as NESB.

Executive Summary

Most research comparing road users from a non-English speaking background (NESB) with those who speak English has examined self-reported attitudes and behaviours, rather than crash involvement. NESB groups have exhibited lower levels of restraint use (particularly child restraint use) in a number of studies. NESB groups may have a lower level of knowledge about the dangers of some behaviours (e.g. speeding and drink driving), but this is not always accompanied by more frequently displaying dangerous behaviour. One Australian study that did assess crash involvement used self-report data and found that NESB women were at a greater risk than their English-speaking peers.

Other variables such as socioeconomic status, education level, employment, residence in areas of high population density, and duration of residence may be at least as important as ethnicity as a predictor of crash involvement. However, many individuals from non-English speaking backgrounds - particularly new arrivals - are often disadvantaged by more than one of these factors as well as having poor English skills. Statistics should also be analysed to take account of exposure, such as unit of travel or number of trips taken.

While most NESB-specific safety campaigns have not been evaluated, there is general agreement that care should be taken in designing the materials rather than simply translating English pamphlets. In designing promotions the reasons underlying any observed differences in the populations should be investigated. Ideally assistance should be sought from the relevant communities in focus-group testing to ensure that materials are appropriate and the meaning of the message is clear. Visual materials should use symbolic or pictographic presentations rather than text to increase their generalisability. Finally, the effectiveness of campaigns should be evaluated to ensure that the message and targeting are efficient and relevant.

As crash statistics databases maintained by the Police do not include an ethnicity variable, the project aimed to examine crash involvement of NESB groups using other datasets. Two separate hospital injury databases - the Victorian Emergency Minimum Dataset (VEMD) and the Victorian Admitted Episodes Dataset (VAED) - code preferred language and country of birth respectively. However, in neither case is the recording of either of these variables sufficiently complete for a meaningful analysis.

Other Australian hospital datasets were investigated, but were not coded any more completely than Victoria's or the number of NESB cases was simply too small for a meaningful analysis. The National Coronial Information System, various comprehensive and third-party insurance databases, and ABS data were also unable to provide useful or complete information to determine the crash involvement of NESB groups.

Generating a more complete set of VEMD or VAED data is the most attractive option, possibly concentrating on a particular hospital or network in a high density NESB location. The time needed to accumulate sufficient data would be dependent on the number of hospitals participating and the percentage of their clientele who can be classified as NESB.

A less rigorous alternative would be to compare the crash statistics between LGAs that have high and low NESB densities. NESB density and composition for individual LGAs is available in Census data. A difficulty with this approach would be matching the LGAs for variables other than ethnicity.

Sponsoring Organisation: Baseline Research Program - Department of Justice, Transport Accident Commission, Royal Automobile Club of Victoria (RACV) Ltd, VicRoads.