Loading...
  • Customer Experience
    Customer Experience
  • Employee Experience
    Employee Experience
  • Brand Experience
    Brand Experience
  • Product Experience
    Product Experience
  • Core XM
    Core XM
  • Design XM
    Design XM

Data & Analysis Basic Overview

What's on This Page:


Was this helpful?


This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

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!


About Data & Analysis

The Data & Analysis tab lets you filter, classify, merge, clean, and statistically analyze your response data:

  1. Click Data & Analysis to reveal up to six key sections.
    image of the data & analysis banner showing all of the subsections
  2. Select between the sections as introduced below:
    • Data
    • Text
    • Stats iQ
    • Predict iQ
    • Crosstabs
    • Weighting
Qtip: Some of these tabs, like Predict iQ and Stats iQ are add on features. If you’d like to purchase access to them, contact your Qualtrics Account Executive.

Data Section

Most filtering, classifying, merging, importing, and data cleaning activities take place under the Data section. For example:

  1. Display and review results in the responses window by toggling between your:
    image of the data & Analysis tab. The Add filter button on the top left is highlighted and the response count in the top right is highlighted

    1. “Completed” or Recorded Responses
    2. “Incomplete” or Responses in Progress
  2. Filter by specific questions or by your saved filters, survey metadata, contact fields, or embedded data fields.
  3. Use operators to drive your filters.
    image of the data tab. a filter is applied to the data. the export & import button is selected.

    Qtip: See the Filtering Responses page for more details.
  4. Navigate page-by-page through your responses.
  5. Click the Export & Import dropdown menu to import responses, combine responses from multiple surveys, manage your previous downloads, or export your data in various formats.
  6. Click the Edit button to enter edit mode; you can add answers to individual responses or make essential edits, such as removing foul language.
    image of the data tab. Editing mode has been activated by clicking the Edit button. The tools dropdown in the top right is expanded

    Qtip: See the Response Editing page for more details.
  7. Delete responses, save column layouts, or translate responses under the Tools dropdown menu.
  8. Click any column header to move, hide, rename, or sort columns (e.g., Largest-Smallest). You can also view responses as numeric values (also known as recode values) in a column.
    image of the data tab. A column of data is selected. The actions column on the far right is selected and the available options are visible.

    Qtip: See the Recorded Responses page for more details.
  9. Use the Actions dropdown menu to delete individual responses, export data to a PDF, or issue a survey retake.
    Qtip: See the Retake Survey Link page for more details.

Text Section (Text iQ)

image of the text tab of the data section

The Text section contains Text iQ tools. Use Text iQ to tag text entry responses with topics for analysis. In the below image:

  • Multiple topics have been tagged in the Topics pane (e.g., “food,” “burgers,” and “ice cream”).
  • Multiple topics can be assigned to your responses.
  • Lemmatization includes various forms of words (such as “burger” and “burgers”).

Text iQ generates various widgets that will give insight to your text analysis. For example, this constellation chart displays the frequency with which certain terms appeared in all the responses. Dots in the constellation get larger as the term appears more frequently.

image of the text tab of the data section with a keyword web visible

Create as many new topics as you need to explore your text responses in depth. Additionally, the Text iQ search will help you pinpoint your topics. Lemmatization and spell check will also speed up your topic tagging:

  • Lemmatization: Dissects words by their roots and tags them accordingly (e.g., car = car, cars, car’s, cars’).
  • Spell check: Variations of misspelled words like “ice creem” or “icecream” will be tagged as “ice cream.”
Qtip: See the Text iQ page for details on how to get the most out of Text iQ.

Stats iQ Section

When you click the Stats iQ section, you’ll open Qualtrics Stats iQ. This will let you dig deep into your analysis, identify trends, and produce predictive models. Stats iQ is a powerhouse statistical tool that can be appreciated by novice and expert analysts alike.

Attention: Stats iQ is an add-on feature not included in the standard Qualtrics license. If you do not have a Stats iQ section, please contact your Qualtrics Account Executive for more information.

Access Stats iQ by selecting the Stats iQ section. From there:

  1. Use the variable pane to select your variable. You can include any, or all, of the questions in a survey. (The search box lets you quickly locate a single variable.)
    image of the stats iq section of the data tab. a workspace with a regression analysis is visible.
  2. Choose one of the analysis buttons (Describe, Relate, Regression, or PivotTable) to trigger statistics for your selected variables.
  3. Each analysis will appear as an analysis card in the workspace with each new card appearing above previous cards. A collection of cards is called a workbook.

Stats iQ can analyze data collected in Qualtrics, as well as any outside data you’ve uploaded following the Imported Data Projects instructions. For more information, visit the Stats iQ Basic Overview page.

Predict iQ Section

Attention: Predict iQ is not available for all users. If you are interested in this feature, please contact your Qualtrics Account Executive to see if you qualify.

Predict iQ analyses your respondents’ survey responses and embedded data in order to predict when a customer will eventually churn (abandon the company). Once a churn prediction model is configured in Predict iQ, newly collected responses will be evaluated for how likely the respondent is to churn, allowing you to be proactive in your company’s customer retention.
image of the predict iq tab of the data & analysis section

For in-depth information about using Predict iQ, check out our dedicated support page!

Crosstabs Section

Crosstabs perform multivariate analysis (i.e., analyzing two or more variables at a time) while calculating p-value, Chi-Square, and T-Test stats.

Typically used with multiple choice and matrix table questions, you can also add embedded data to your crosstabs.

In any crosstab visualization, rows are called stubs and columns are called banners.

Cross Tabs, a gray matrix with a blue top row of labels, and percentages and capital letters

In the above image, you can see that marital status is being associated with level of education.

The capital letters in this image represent statistical comparisons between columns. In the case of where the “Divorced” column intersects with the “Bachelor’s degree” row,  there is a capital C. This means that respondents that are Divorced are significantly more likely than those in column C, the Never Married respondents, to indicate that they have a Bachelor’s degree.

Qtip: The example above was an example of the Column Percentages (Answering) and Column Stats Test (Answering) calculations. Review the Crosstabs page for step-by-step instructions on how to perform these kinds of calculations, and more

Significance is determined by adjusting the Confidence Level. Click the button in the upper-right to change the threshold in which difference are considered statistically significant.

After clicking the confidence level button in the upper-right, a new window opens overtop with a dropdown where a new confidence level can be selected

Qtip: Read the Crosstab Options page for instructions on additional options that can inform your calculations.

Weighting Single Variables

Surveys sample larger populations. The Weighting section lets you adjust your sample to account for underrepresented populations.

For example, say that you are surveying conference attendees from all over Canada, yet you need your sample to proportionally reflect each province. You may want to re-weight the survey to reflect your desired distribution. In the below image:

  • Alberta is underrepresented hosting just 7.5% of survey takers, but 11.57% of the population of Canada.
  • Ontario is also underrepresented containing 38.26% of the population, yet only contributing 7% of the survey respondents.
  • The Yukon territory is overrepresented with 10.4% of the survey participants, yet they are only .10% of the Canadian population.

Example of weighting responses

It’s easy to change the weighting (by percentage) in the Target weighted column. Your percentages will equal 100% in the end.

Response weighting can be applied to your Results-Reports. You can turn your weightings on or off globally (for a whole report) or for a single visualization (a graph or table).
"Use weighted metrics" option in the Results-Reports Global Options

Qtip: For step-by-step instructions on how to apply weighting, visit the Response Weighting page.

Weighting Multiple Variables

There are two options for weighting multiple variables:

  • Raked Weighting: Calculate two (or more) variables independently and display them side-by-side.
  • Interlocked Weighting: Overlay two or more demographics. For example, melding a “Years of Experience” variable with a “Province” variable.
Qtip: There is a lot that can be learned with interlocking weighting. For example, there is a noticeable experience gap between participants from Alberta versus Ontario in the above image.
Qtip: For more information about interlocked and raked weightings, explore the Response Weighting page.

Making Decisions About Weighting

There are many potential causes of a skewed survey:

  • Non-response: Certain demographics fail to respond to your survey.
  • Panel design: Your panel (list of targeted respondents) has not been properly selected.
  • Self-selection: Those who opt into your survey do not reflect your targeted demographic.
  • Sample size: You may not have enough respondents to be significant.

This is where the Weighting section is helpful. It’s here that you can study and interpret problems with the representativity of your data that may make it difficult to reach reliable conclusions based upon your response distribution.

You may learn that you need to:

  1. Reopen a survey and target additional respondents from the underrepresented population, or
  2. Change the weighting of the existing responses to better reflect the integrity of your research goals.
Qtip: Read more about interpreting statistics on the Understanding Statistics page.

FAQs