Assigning Weights on a Question and on a Dataset Level
The user can nominate a question to be a weight and then specify that this weight is “default” for a dataset. In this case, all results for this very dataset will be by default weighted by this new weight, instead of the primary weight for the entire study family.
On a more specific level, the user can choose to assign a dedicated weight only to a single question. Then this very question results will be by default weighted by the new weight, and not by the primary weight for the entire study family nor the default weight for the dataset.
To define a weight on a dataset level
Go to the edit mode and use a question as an alternative weight. Only the alternative weight can be assigned to a dataset (as the primary weight will be used by it anyway). When the alternative weight is selected, look for the “Default for this dataset” switcher under the “Weight name” in the 2nd column. If this switcher is set to “On”, this very weight will be the default one, meaning the results for this dataset will be weighed by it, unless specified otherwise in the “Calculation” section in the context menu in Query.
Only a single weight can be selected as default for a single dataset (although each dataset in the study family can have a different default weight). If you already specified the default weight and then go to the other alternative weight within the same dataset, the described switcher will be unavailable (with red stop sign instead of a cursor) and a proper description will be provided in the infocloud.
To define a weight for a specific question only, select that question in the Data Import. In the 2nd column, under the “Question type” field, there will be an “Assigned weight” field. In this drop-down field, few entries will be available:
-
-
List of actual weights specified for this study family and available in this dataset (including the primary weight).
The first one is selected by default for each question in the dataset. When selected, the question is weighted in accordance with the dataset weight - primary, or the default one specified for the dataset. If the user selects any specific weight from the list, this question will be weighted by it, and the choice is static.
Examples:
-
The user specified exactly that question X should be weighted by the primary weight. After that they select weight A as a default for the dataset. All questions apart from question X will be then (by default) weighted by weight A. As question X has primary weight assigned specifically by name, this question will be still weighted by the primary.
-
The user selects weight A as default for the dataset and then assigned weight A to question X. After that, they change the default weight for the dataset to weight B. All questions apart from the X will be weighted by B, but the question X will be still weighted by A.
-
The user selects weight A as a default and keeps “As for the dataset” for question X. After that, they deselected weight A as a default, meaning the entire dataset, including question X, will be weighted by primary.
Review questions with assigned weight
If you assign a weight to a question, an additional tile will appear within this question icon with the W letter. Also, if you select the weight and look to the 2nd column, a section “Assigned weight for” will be available to review in how many cases this weight was assigned to a question.
If you downgrade the weight to a normal numeric question, the error icon will appear next to the questions which have this weight assigned and you will see a red text “Missing weight” in their “Assigned weight” field.
