Troubleshooting weight schemes

Refer to this topic if you are encountering errors when you try to create and apply a weight scheme.

Target percentages

Check the target percentages in your weight scheme and ensure they meet all of the following criteria.

Valid target percentages must:

  • Add up to 100% total.
  • Contain numbers only (not letters or special characters).
  • Be positive numbers (not negative numbers).
  • Be set to 0 if the sample percentage value is 0.
  • Be set to a value above 0 if a value above 0 exists for the sample percentage.

Weight variables

Error How to fix it
The weight variable does not have answer options. Choose a weight variable with 2 or more answer options.
The weight variable has 1 answer option. Choose a weight variable with 2 or more answer options.
The weight variable has no responses. Choose a weight variable that has responses.
The weight variable was deleted. Remove the weight variable from your weight scheme.
The weight variable changed (for example, answer options were added and removed). Remove the weight variable from your weight scheme and re-add it.
The weight scheme is invalid due to sample percentage changes (for example, receiving more survey responses or report filtering).
  • If a sample percentage value is now 0, change the target percentage value to 0.
  • If a sample percentage value is now above 0, change the target percentage value to something above 0.
The weight scheme fails because multiple variables contradict each other and the targets cannot be reached.
  • Remove variables from the weight scheme.
  • Adjust the targets so that they are closer to the original sample.
The weight scheme fails because a weight variable in a multi-variable weight scheme has missing values for some participants. For example, participants skipped a survey question, or there are missing profile variable values. Filter the report to exclude participants with missing values. For profile variables used as weight variables, create a filter condition statement that includes all profile variable values. This will implicitly exclude participants who do not have values for that profile variable.
The weight scheme fails because there are outliers or disproportionately small groups in the weight variables.
  • Adjust your weights to more closely reflect the general population.
  • Merge the outliers with the next biggest group. For example, if you have 5 participants aged 80+ and 100 participants aged 65+, merge the 80+ group with the 65+ group.