Participating households are chosen from a number of randomly selected Census meshblocks. Each selected household is allocated two consecutive travel days to be surveyed.
Over a seven to eight year cycle, every household in the selected meshblock will be invited to participate in the survey. Once this happens, new meshblocks will be selected for sampling.
Some time before the designated travel days, the household is sent an introductory letter outlining the survey and advising that a surveyor will call. During the week prior to the travel days, an interviewer visits the home, and with the agreement of residents explains the survey (Pamphlet about survey (PDF, 1,903 KB) and leaves Memory jogger (PDF 143KB) in which respondents can note their travel on the selected days.
As soon as possible after the travel days, the interviewer returns and conducts a personal interview with each household member, during which his/ her travel details, accident history, alcohol consumption and other personal data are recorded.
Each year approximately 2200 households in 280 meshblocks are invited to take part in the survey.
Survey data are entered into a custom-designed Access database using a laptop computer, at the time of interview. Paper questionnaires are provided for situations where use of the computerised form is not practicable, and the results are entered into the database later by the interviewer.
Data validation checks (including on the spot verification of street names) are incorporated into the database, giving rapid feedback and enabling early correction of errors.
Trip distances for cycle and motor vehicle trips are calculated from the origin and destination addresses provided by the respondents. Critchlow Associates Ltd were contracted to generate automated map coordinates for each address and then to calculate distance based on the fastest route between the origin and destination addresses, or via a specified intermediate address if one was provided.
On Friday, Saturday and Sunday nights, two out of every three fatal crashes involve an alcohol-affected driver