Annual

The data collection is conducted from march to june. With some exceptions, the interview is usually conducted in the 2 weeks and 2 days following the reference week.

Household

The sampling frame is an address database which is updated with an annual cartographic survey on one-fifth of the territory. Mayotte territory blocks have been divided into 5 rotational groups based on 2012 census counts. These rotational groups are stratified by the geographic clustering of all municipalities and are balanced on the number of dwellings and others sociodemographics variables.

The sample is stratified on the size of the address (single dwelling, less than ten dwellings, more than ten dwellings). The sample is balanced on the region. For each region, the sample is balanced, proportionnaly of the number of main residences, on some variables from the sample frame. 

It is a rotative sampling : one third of the sample is refreshed every year.

The sample is uniformely distributed among the 17 reference weeks.

The average sampling rate is 1/66.

Each year, around 3 000 dwellings are surveyed.

The weights are first adjusted by estimating the response probabilities using a logistic regression (with auxiliary variables : type of address, type of housing, geographical location, Wuambushu zone indicator).
Then, a calibration is done on individual margins (disaggregated by gender and age group, place of birth) and on housing external margins (number of main residences). In order to control the collection bias, three other margins were added this year : the number of single housing, permanent buildings and the number of housings in a priority district.