Stay up-to-date with every new feature in Qualtrics XM

LEARN MORE

Ready to try Qualtrics?

GET STARTED

Editing a Data Model (CX)

Suite

Customer Experience Strategy & Research

Product

Qualtrics

What's on this page

About Editing a Data Model

Once you’ve created a data model, you may need to make edits to it. The edits covered on this support page include:

Removing elements.
Changing the fields you can map in your dataset.
Managing labels.

If you want to map multiple sources together or join data, see Unions and Joins. For how to create or use a field group, see Field Groups (CX). For recoding values, see Recoding Data Model Fields.

Qtip: A. “node” refers to any element in a data model. This includes sources, unions, joins, and output datasets.

Editing a Data Model

Go to the Data page.

clicking hamburger menu top of every Qualtrics page, selecting data from the menu

Click the 3 dot menu next to the dataset you want to edit.

page displays a list of datasets with 3 dots to the far right of each, which opens a menu once clicked

Click Edit.

Qtip: If your data model has been added to a dashboard, you can also edit it right from that dashboard. Open your dashboard settings, go to Dashboard data, and click the name of the dataset you want to edit. Often, there may be only 1 dataset in your dashboard.

Removing Nodes from the Data Model

Warning: Be careful when editing a data model that’s currently in use. If you delete a node (union, join, or source), then add the same node back, it’s treated as a brand new, separate node. That means you will have to readjust any widgets that use fields that were deleted and re-added.

To remove a source, a union, or a join, click the X in the upper-right.

every source, union, join, and output appears as a separate block in the model. X in upper-right of the rightmost block in the model

Selecting Which Fields To Include from Sources

If you click a source and go to Schema, you can select and deselect the fields included in your dataset.

clicking a source block on the left, which opens a menu along the bottom of the page; in this menu, you can go to the source tab

Qtip: The steps described here control the fields you can possibly map in your dataset. To add a new field, see Unions.

Adding a New Field

There are many reasons you might need to add a new field to your data model. For example, if you added a new field (or column) of data to your source after you published your data model, or if you changed your mind about a field you didn’t map earlier.

You can add a new field to many nodes, including:

Data sources
Unions
Joins

However, adding a field to one node will not automatically carry it into any other connecting nodes. You must add the same field to its source and to each following node to include it in your dataset output.

Example: You added an Education field to your CSAT 2019 survey. To get Education into your final dataset, you need to first add it to your CSAT 2019 source, then in Union 2 and the Join. The field doesn’t need to be added to Union 1 because it isn’t connected to the CSAT 2019 survey.

If you’re editing an existing dataset, remove the output dataset.

Select the data source, union, or join that you’d like to add a field to.

Adding the Education field to the source

Go to the Field editor.

Click Actions.

Select Add field.

Map and name the dashboard field.

Select the field type.

Add the same field to any following unions or joins that are connected.

Attention: When mapping the same field across multiple nodes, be careful to use a consistent field type. For example, if your CSAT is a Number Set at the source, it should be a Number Set in any unions and joins that follow. If you change the type in one node, you must change it across all of the others. If the type differs for each version of the field, there could be unintended effects, such as certain values not carrying over.

Once you’re finished making changes, add a new output dataset at the end.

Publish your data model.

Qtip: Rather than mapping a dashboard field to a source field, you can also map to a static value that will be the same for every row of data in the data source. First, click Create a custom field.

Add your static text in the window that opens.

To learn more, see Static Text.

Deleting Fields

Attention: Make sure the field is deleted from every node in the data model. Deleting the field from the source removes the data from the dataset, but it will leave an empty field in any unions or joins that use that source.

Click the source that contains the field you want to delete.

Go to Field editor.

Click the 3 dots next to the field.

Select Delete Field.

Select any following unions or joins.

Delete the field from those nodes, too.

If you’re editing an existing dataset, make sure to delete the output dataset.

Add a new output dataset.

Publish your changes.

Editing Dataset Fields

There are many changes you can make to your dataset fields. Some edits are made to the source, some are made in unions, and others are made to the output dataset.

To edit a source, union, join, or output, first click on it.

clicking a join, which opens a menu along the bottom of the data model

Editing source fields

Editing unions

See Unions (CX) for specific steps on:

Mapping the same field from multiple sources together.
Separating fields.

Editing joins

With joins, you can merge data rows based on a common field, such as a unique ID. See Joins (CX) for more details.

Editing output datasets

Field Groups (CX)

Data Model Drafts

A draft is created whenever you edit an existing data model and leave the page without publishing. Your data model will have a “Draft” label added to the top of the page.

the "draft" label at the top of the data model -- the "reset current draft" and "publish" buttons are to the right

You can click Reset Current Draft to delete the current draft and revert the model to the last published version.

If you’d like to save your changes instead, click Publish. Publishing applies the data model to the underlying data, so if a draft exists, then the model used in your dashboard may be outdated until you publish changes.

FAQs

How many columns / unique fields can my dataset have?

The maximum number of unique fields (i.e., columns of data) is 2,000 in each dataset.

Is there a limit to the number of data sources you can add in each dataset?

You can only have 65 data sources in each dataset.

How many dashboard datasets can I create?

See Dataset Utilization Limits.

How many times can I use the same data source in a dashboard dataset?

See Dataset Utilization Limits.

Are changes to data models reflected immediately in dashboards?

It depends. Left outer joins and unions are both continuously updated. However, some other join types update periodically. See Continuous vs. Periodic Dataset Updates.

If you have multiple sources of the same type in your dataset (such as tickets and surveys), we generally recommend creating unions before you create joins.

What’s the difference between data sources and datasets?

Data sources include specific places the data comes from, such as surveys, tickets, directories, and more. Datasets are collections of multiple data sources where multiple fields are either mapped together (also called a union) or joined together.

Learn more about these key terms.

Was this helpful?

The feedback you submit here is used only to help improve this page.

That's great! Thank you for your feedback!

Thank you for your feedback!