ENUMERATION

The enumeration was done using the direct interview or canvasser method. The counting was de facto, that is, people were enumerated in the place where they passed the night of the census. The de jure population was also counted and included the population who usually live in the dwelling regardless of where they passed the night of the census. Only five items were included in the questionnaire on usual residents absent during the census night (age, sex, relationship to head of household, location on census day and duration of absence).

The census interviews were conducted for approximately three weeks after the census commencement date (11 July 2004). The census questionnaires had to be answered according to respondent's location and circumstances on the date designated as the official census date, that is, 11 July 2004.

As explained above, there were special questionnaires for institutions and hotels. Special interviewers were appointed to collect data in those establishments.

District Managers transferred waypoints from GPS receivers to computers. Each receiver was numbered. Downloaded files were named with that number so it was always known which GPS was used and where. Downloaded location files typically contained three fields: 6-digit waypoint number, latitude, and longitude. Content files included dwelling numbers (from the detachable part of the household sticker) and the coded content of the questionnaires. Each GPS waypoint record was linked with its corresponding questionnaire record. 

In general, all GPS operators knew their job. However, in a few instances, some operators tried to be too clever and they entered waypoints at home, at the kitchen table, working through their list of household sticker numbers. Obviously, they were found out the moment the set of waypoints of a particular GPS was downloaded and the points showed up in a tight cluster over one house on the aerial photograph.

As the results were coming in, questions arose regarding areas that had been missed. A team consisting of mapping staff and Census Team Leaders from Dili was standing by to quickly cover areas where GPS activities appeared to have gone wrong (see Box 5).

Because of lack of resources, a post-enumeration survey was not conducted. It is known that the 2004 Census did not have 100% coverage of dwellings and households, but it is considered that the coverage was well over 90%. Timor-Leste is very proud to claim to be the first country anywhere in the World to base its national census on a GPS enumeration of almost 100% of its households.

The sub-district census offices closed by 14 August, 5 weeks after census day, and district offices closed a week later. The last task to be done after completion of the interviews was to write up the provisional counts of males and females on questionnaire covers. These counts provided the early provisional estimate of the total population for Timor-Leste which was to be announced by the Prime Minister. The field work was finished on the weekend of 21/22 August, when district managers came back into Dili with all their questionnaires and remaining materials, ready to start their field debrief.