Documentation Note : merge with http://analysishelp.askia.com/weighting_analyse

Definition

The process of weighting involves changing the weight of the interviews to correct sampling errors.

Here is an example: we interviewed 60 men and 40 women. The population we want to describe is equally composed of men and women. To correct this, each time we run counts on a question, instead of adding 1 for each man, we will add 50/60 = 0.8333, and for each woman we will add 1.25. In other words, the under represented categories will have a stronger vote than the over represented ones.

Multiple targets

The process gets more complicated when more than one variable is involved. Let us imagine we want to apply a weighting on gender and social grade (A,B,C1). If we have the information on how the gender is defined for each social grade (and hence creating a new variable with 6 cells), we can fall back on the one variable weighting. But if we increase the number of questions and responses in the weighting, the target population frequency in each of the sub-cells may not be known.
Another alternative would be to assume that gender and social grade are independent and therefore we should observe on each social grade a 50-50 split on gender. This hypothesis is very restrictive and would very likely skew the results. Therefore the algorithm is an iterative process.

1. First the weights are set to 1 for all interviews.
2. The system finds which variable is the furthest from the targets.
3. The current weights are changed so this variable will exactly fit the targets.
4. The minimum and maximum weights are applied if they are defined.
5. If all variables fit the targets then the process will stop otherwise it will start again at b).

After a few iterations the weights should converge to an acceptable solution.