What does re-allocate weights to available variables mean?

Sometimes data in a score is missing. When you define a score, you can specify how the score should be calculated when data is not available.

The most commonly used method for handling missing data is to re-allocate weights to the variables that do have data. The weight of the variable missing data is distributed to the other variables in the score, in proportion to each variable's weight.

This concept is best illustrated with an example. Let's say you define an application score called "Application Score" which weights a credit applicant's PAYDEX® Score at 40%, Commercial Credit Score Percentile at 30%, and Years in Business at 30%.

Application Score

Variable

Weight

PAYDEX® Score

40%

CCS Percentile

30%

Years in Business

30%

Total Weight

100%

If Years in Business is not available, its 30% weight is distributed to the other variables in the score in proportion each variable�s weight.

Variable

Weight

Reallocated Weight * (see note)

PAYDEX® Score

40%

57.14%

CCS Percentile

30%

42.86%

Years in Business

30%

0%

Total Weight

100%

100%

In this example, notice that the reallocated weights add up to 100% and the ratio of the PAYDEX® Score weight to the CCS Percentile weight is the same both with and without reallocation (57.14 to 42.86 ratio with reallocation = 4 to 3 ratio without reallocation)�

* Note: The reallocated weight is calculated as:

Variable Weight + Variable Weight / (100 � Missing Variable Weight) x Missing Variable Weight

For example, the reallocated weight for PAYDEX® Score is calculated as:

40 + 40 / (100-30) x 30 = 57.14%��