Dapresy supports standard target (cell) weighting and RIM (Random Iterative Method).

**Standard target weighting:** cell weighting is based on the matrix of the variables you want the sample to be representative for. For example if you would like to weigh your data to age and gender and you know how those groups interfere with each other (grey area in the table below), you could weight the cells:

Example of representative numbers of gender age:

gender/age | < 35 | >=35 | total |

men | 0.30 | 0.31 | 0.61 |

women | 0.28 | 0.11 | 0.39 |

total | 0.58 | 0.42 | 1 |

Weights can be calculated by dividing the population (P) by sample (S): P/S. Here you can find some examples:

## Example 1:

A Gender variable is selected to be the base in the weight calculation. The variable contains the answer options Male and Female. The following distribution is entered:

- Male: 0,51
- Female: 0,49
- “Full period” is the selected time interval.

The imported data contains 900 respondents: 300 Female and 600 male respondents.

The weight number for each Male:

Since the male distribution is set to 0,51 it means that the weight of all male respondents should sum up to 51% of 900. The sum of all males should be 0,51*900 = 459. The weight number of each Male respondent should be 459 (P) /600 (S) = 0,765.

The weight number for each Female:

Since the female distribution is set to 0,49 it means that the weight of all female respondents should sum up to 49% of 900. The sum of all females should be 0,49 * 900 = 441. The weight number of each female respondent becomes 441 (P) /300 (S) = 1,47.

## Example 2:

A Gender variable is selected to be the base in the weight calculation. The variable contains the answer options Male and Female. The following distribution is entered:

- Male: 0,49
- Female: 0,51
- “Month” is the selected time interval.

The imported data contains:

- 900 respondents (300 Female and 600 male respondents) in January
- 420 respondents (200 Females and 220 males) in February.

__The weight number ____for January respondents__Weight number of each female respondent

Population should be 0,51 * 900 = 459. In the sample we have 300 women. So weight value will be 459 / 300 = 1,53

Weight number of each male respondent

Population should be 0,49 * 900 = 441. In the sample we have 600 men. So weight value will be 441 / 600 = 0,735

__The weight number ____for February respondents__Weight number of each female respondent:

Population should be 0,51 * 420 = 214,2. In the sample we have 200 women. So weight value will be 214,2 / 200 = 1,071

Weight number of each male respondent

Population should be 0,49 * 420 = 205,8. In the sample we have 220 men. So weight value will be 205,8 / 220 = 0,9354545455

## Example 3:

A Gender variable is selected to be the base in the weight calculation. The variable contains the answer options Male and Female. The following distribution is entered:

- Male: 0,49
- Female: 0,51
- “Full period” is the selected time interval.

The imported data contains 910 respondents: 300 Female and 600 male respondents. 10 respondents have no Gender markup.

Weight number of each female respondent:

Population should be 0,51 * 900 = 459. In the sample we have 300 women. So weight value will be 459 / 300 = 1,53

Weight number of each Male respondent:

Population should be 0,49 * 900 = 441. In the sample we have 600 men. So weight value will be 441 / 600 = 0.735

*NOTE: Weight number for the respondents without Gender mark up, the system always sets the weight number to 1 in the calculations if the weight is missing (counted as 1 in calculations).*