The proposed methodology involved dividing the township into some 11 sectors, marked on a map of the township provided by the Aurukun Shire Council. A total of 134 dwellings had been identified from the map for Indigenous residents. The map of the township was to be used to develop a preliminary Dwelling Check List for each sector. Those of non-Indigenous (primarily staff) residents, for which the standard census forms were used, were additional, and the CC undertook the responsibility of providing and collecting these standard census forms.

Responsibility for filling in and collecting the forms for the Indigenous households in each sector was to be assigned to one of five two-person teams of interviewers. The CC ensured that each team had at least one person who was fluent in or had a good working knowledge of Wik Mungkan, the lingua franca of the Aboriginal residents of the area, and at least one person who was competent to fill in the SIHFs and SIPFs accurately. In some instances, this meant that teams comprised a Wik and a non-Wik person, while in others one Wik person involved satisfied both criteria and was assisted by another Wik interviewer.

In assigning interviewers to teams, the CC tried to ensure that the capacity of each team to operate effectively was not compromised by personal incompatibilities between team members (such as those arising from kin-based restrictions for Wik interviewers), or by potential incompatibilities between interviewers and the residents of their assigned collection sectors. An important source of such potential problems was the high level of disputation within Aurukun, structured in terms of residence in the 'topside' (eastern) and 'bottomside' (western) areas of the township. As discussed elsewhere (Martin 1993), this basic division within the township reflects long-standing patterns of political and social relations between 'topside' or inland Wik groups and 'bottomside' or coastal ones.

The census collection revolved around three key individuals; the (non-Aboriginal) CC, whose main role centred on organising the logistics and ensuring the rigour of the collection, and two Wik individuals, one male and one female, who between them had an almost encyclopaedic knowledge of kin and family relationships and of household compositions across the township. Particularly in the case of the male Wik interviewer, the knowledge gained from being a member of the Wik community resident in Aurukun had been supplemented by his long-term role within the Shire Council administration. This meant that he had an unparalleled knowledge of the kinds of information which would normally be held in administrative data sets, such as household compositions, sources of income, income levels, information on employer details, and so forth.

The census collection operation was based in the large and secure air conditioned restaurant at the rear of the Council-run tavern. Tables had been set up around the room where paperwork for each sector was kept, including a copy of the relevant section of - 17 - the township map and ultimately the collated SIHFs and SIPFs. In the hot and dusty conditions in Aurukun, having such a facility provided an invaluable aid in maintaining a systematic and ordered approach to the collection and its associated paperwork.

It was proposed that the names of those who were normally residents of Aurukun but were absent during the week of the census (e.g. at the Pormpuraaw football carnival) would be placed on the relevant SIHF. Largely blank SIPFs would also be created for them, and interviewers would attempt to get the relevant information over the two or so weeks that it was envisaged the census collection would take to complete. The strategy was clearly based on the assumption that absent usual residents would not be enumerated elsewhere, which at least in the case of those at Pormpuraaw was certainly justified. Visitors in Aurukun households were also to be enumerated.

It was also proposed that once all the SIHFs and SIPFs had been collected, an information validation process would be conducted, involving the CC and the principal Aboriginal interviewer. It was intended that this exercise would focus particularly on ensuring that children had been enumerated, working from the Aboriginal collectors' knowledge of who were care-givers and which children they looked after. Another proposed aspect of the validation process was ensuring that there had been no double-counting, for example for people who moved between two or more residences on a regular basis. The final stage was to involve 'spot checking', at places such as the community store, to estimate how thoroughly the census had been done.