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

XM Essentials 9: Filtering Out & Weighing In

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!

Filtering Data

Filters are logic-driven statements that can make editing easier. In our example, many respondents spelled “root beer” as a single word (“rootbeer”). But, there are too many records to go through to reliably find each misspelling one-by-one.

Try it!

Exercise A-F: Filtering to Spot Errors

  1. Open your Project.
    Selecting our project from the projects page
  2. Choose the Data & Analysis tab.
    Tabs along top of a survey
  3. Click Data.
  4. Click on the Add Filter button.
    Filter button upper-left of Data and Analysis tab
  5. Choose the Text – Other part of this question: What summertime drinks do you prefer?
  6. In the second dropdown menu under Select Operator, choose Contains from the list of operators.
    Filling out dropdowns of a filter
  7. In the third dropdown menu, enter root.
    Qtip: With this filter in place, all the records containing root, root beer, and rootbeer will appear.
  8. Click Edit.
  9. Fix all the spelling errors.
  10. When finished, remove your filter by clicking the minus sign.

How Our Text Data Improved

Our word cloud scrubbed up nicely and now communicates the results more clearly without misspellings and expletives. (Compare this word cloud with the unedited one we showed you in XM Essentials 8.)

A green word cloud. There are no weird fillers or spelling errors

Filtering for Analysis

Filters are powerful tools to help you gather insights from your responses. Don’t limit yourself to just a few filters or you may miss something. For example, demographic differences and price sensitivities are often key variables to filter. In this example, we will filter parents from teenagers and older patrons, as well as how they answered regarding the price they are willing to pay for drinks.

Try it!

Exercise A-G: Filtering for Demographic Differences

  1. Open your project.
    Selecting our project from the projects page
  2. Choose the Data & Analysis tab.
    Tabs along top of a survey
  3. Click Data.
  4. Click on the Add Filter button.
    Add filter button in upper-left
  5. Choose the question, “How old are you?” from the dropdown list. (If you are using a different survey, select a similar question type to filter.)
  6. In the second dropdown menu under Select Operator, choose Is from the list.
    Dropdowns in a filter
  7. In the third dropdown menu, choose the demographic age ranges of your potential neighborhood parents (25-34, 35-44, and 45-54).
    Last dropdown of a filter contains the answers from the question you are filtering by

    Qtip: With this filter in place, only response records from participants ages 25 – 54 can be seen.
  8. Remove your filter before creating another by clicking the minus sign ( ).
  9. Repeat the steps above and practice filtering out everyone under age 11 and over the age of 24.


It turns out that more young people answered Moksh and Naman’s survey than their parents. Also, while some older neighbors have taken the survey, few of them have braved the summer heat and made a trip down to the Lemonade & Lassi stand.

In survey terminology, both young people and the elderly have been oversampled. Conversely, parents are under-sampled. When we ran a Results-Report, here’s how the demographics broke out.

Red horizontal bar chart

Response weighting lets Moksh and Naman balance the responses by age.

Try it!

Exercise A-H: Weighting Responses

  1. Go to the Data & Analysis tab.
    Picture of the weighting tab with the create new scheme button in the center
  2. Click Weighting.
  3. Click Create new scheme.
  4. Choose the type of weighting you’d like to try. We’re going to use cell-based because it’s the easiest; you just need to make sure all your groups add up to 100%!
    Three weighting options, with cell-based highlighted
  5. Click on the Add a variable button.
    Add weighting button expanded to show list of questions from the survey
  6. Choose the question “How old are you?” from the dropdown list. If you are using a different survey, pick a similar question type.
  7. Notice that children under 10 are 32.6% of the respondents, but the big spenders are the adults. Parents are also likely to be big influencers, driving to a degree the purchasing habits of 1-10 year olds.
    Weighting page
  8. Do a simple weighting. Here, we balanced all the groups 10-44 to 15%. Then we set the 45-54 group at 10%, and gave everyone 55 and up 0%, due to the observation that the older demographic group rarely, if ever, ventured to the lemonade stand.
    Qtip: Your percentages must equal 100%.
  9. Click Next.
  10. Click the Save Weights button.
    Picture of the save weights button

Qtip: Here’s a sneak peek of how the weighting dramatically changed the bar graph in our results.

Horizontal bar chart

Qtip: Crosstabs and Stats iQ are not compatible with response weighting. Quotas provide a more reliable way to curb oversampling with these features. Learn more about Response Weighting and Quotas on their respective support pages.

What’s Next?

  1. Continue to XM Essentials 10: Report to the Neighborhood.