Desktop User Guides > Professional > Data management scripting > Working with the Weight component > Rim weighting formulae
 
Rim weighting formulae
Root mean square
The formula is given for a rim weighting matrix consisting of two variables (dimensions), but the same principle applies when there are more dimensions.
Notation
Represents
This graphic is described in the surrounding text.
The target number in category i in the first dimension
This graphic is described in the surrounding text.
The sum of the observed numbers in category i in the first dimension
This graphic is described in the surrounding text.
The target number in category j in the second dimension
This graphic is described in the surrounding text.
The sum of the observed numbers in category j in the second dimension
This graphic is described in the surrounding text.
The number of categories in the first dimension
This graphic is described in the surrounding text.
The number of categories in the second dimension
The formula for the weight adjustment is
This graphic is described in the surrounding text.
Where
This graphic is described in the surrounding text.
Represents the weight adjustment calculated in the previous iteration for the cell at the intersection of category i in the first dimension and category j in the second dimension. In the first iteration, it is substituted with
This graphic is described in the surrounding text.
This graphic is described in the surrounding text.
Represents the sum of the weight adjustments calculated in the previous iteration for category j in the second dimension. In the first iteration, the expression
This graphic is described in the surrounding text.
is substituted with 1.
The calculation for the root mean square is:
This graphic is described in the surrounding text.
where
This graphic is described in the surrounding text.
Represents the sum of the weight adjustments calculated in the previous iteration for category i in the first dimension.
This graphic is described in the surrounding text.
Represents the sum of the weight adjustments calculated in the current iteration for category j in the second dimension.
At the end of each iteration, the Weight component compares the root mean square with the product of the weighted total and the given limit. (The limit defaults to 0.005, but it can be set to another proportion.) The iterations continue until all of the weights are within the limit or the maximum number of iterations has been reached.
Rim weighting efficiency
The rim weighting efficiency figure gives an indication of how well balanced the sample is.
Let
This graphic is described in the surrounding text.
Be the preweight for case j
This graphic is described in the surrounding text.
Be the rim weight for case j
Then the rim weighting efficiency is:
This graphic is described in the surrounding text.
If the data for many respondents needs to be weighted heavily up or down, the efficiency percentage will be low. The greater the percentage, the more well balanced the sample.
Further information
For more information on rim weighting, see the Rim Weighting Theoretical Basis Paper entitled “ON A LEAST SQUARES ADJUSTMENT OF A SAMPLED FREQUENCY TABLE WHEN THE EXPECTED MARGINAL TOTALS ARE KNOWN”, by W. Edwards Deming and Frederick F. Stephan, in Volume 11, 1940 of the Annals of Mathematical Statistics.
See
Working with the Weight component