Rim weighting

For rim weighting, the weighting report shows the target value for each category of each variable on which the weighting is based. In addition it includes additional information that is not relevant to target and factor weighting.

▪Input frequency. The number of cases for which the category is selected.

▪Input percent. The percentage of cases for which the category is selected.

▪Projected frequency. The number of cases that are required in the category. This is based on the target given for the category.

▪Projected percentage. The percentage of cases that are required in the category. This is based on the target given for the category.

▪Convergence/failure. The iteration on which convergence occurred or on which the weighting was deemed a failure.

▪Root mean square (RMS). Rim weighting calculates the weights for each cell in the weighting matrix using a form of regression analysis. The aim is to distort each variable as little as possible while trying to attain all of the desired proportions among the various characteristics. The root mean square figure tells you how much distortion has been introduced, and therefore how reliable your sample is. The larger the figure, the greater the distortion and therefore the less accurate your sample. For details of the formula, see Rim weighting formulae.

▪Limit. This is the limit below which the root mean square must not fall. It is calculated as the product of the weighted total and the limit defined in the Weight.RimLimit property (0.005 by default).

By default, the following information is given for each category of each variable in the final iteration. However, you can request this information for each iteration by using the Weight.RimReportInDetail property.

▪Rim weights. The rim weight calculated for the cases in the category. Weights that are outside of the required range are highlighted. The range is defined in the Weight.RimFlagBelow and Weight.RimFlagAbove properties, which are set to 0.6 and 1.4 by default.

▪Output frequency. The number of cases in the category after the weighting is applied.

▪Output percent. The percentage of cases in the category after the weighting is applied. Rim weighting runs an extra iteration to calculate the error for the final iteration. Because of this, the Output percent will not match the target that is shown as Given in the report.

▪Rim weighting efficiency. This figure gives an indication of how well balanced the sample is. If the data for many cases needs to be weighted heavily up or down, the efficiency percentage is low. The greater the percentage, the more well balanced the sample. For details of the formula, see Rim weighting formulae.