Notes
GENERAL NOTES
----------------------
The household weight can correct just for sample bias, or also for non-sampling errors (such as response bias and adjustment to population distribution). In many cases, the weight is a population weight, which means that the weight will inflate the result to reflect the total population.
We strongly recommend that all statistical analysis be performed using the data weights
COUNTRY SPECIFIC NOTES
----------------------------------
Two weight variables were provided in the raw data; one in the household data file (weight) and the other in the expenditure data file (hhweight). Both were included in the harmonized data file and named HHWEIGHT and HHWEIGHT2, respectively. The survey documentations indicates that the 2 variables are identically weighing the sample, and one of them should always be used as a weighing coefficient since the collected sample is not self-weighted.