Skip to content
English
Contact us
  • There are no suggestions because the search field is empty.

Power BI: Data Structure

The Data Structure page is your starting point every time you open your Power BI dashboard. Think of it as a control centre: before you look at any analysis, this is where you configure the filters that will carry through to every other page in the report. Any selections you make here, your date range, which data sources you want to include, and how you want to view your data, will stay active as you move between pages, so you don't have to reset them each time.

How to use this page

Setting your filters

At the top left, you'll find your universal filters. Start by selecting your date range using the date picker. This sets the time period for all of your analysis and automatically applies across every page you visit. This is especially useful if your project holds years of historic data but you only need to look at a recent window, set it once here and you're good to go.

Next, you'll see the Volume Type filter. This lets you choose between viewing data at the author level (counting individual people who responded) or the verbatim level, which we call item (counting each individual piece of feedback). For most qualitative analysis, we recommend selecting item, this ensures every comment is counted, including when one person has left multiple responses.

You can also choose whether to include or exclude blank responses. Blank responses are submissions where someone has left a score but no text feedback. These are not assigned a sentiment score in Power BI. If you want to see scores only from people who also left a comment, select Exclude Blanks. If you want all responses (including those who only rated), select All.

For more detail on how these choices affect your analysis, refer to the Key Points section below. As a general rule, we'd recommend starting with item level and including blanks, and only deviating when there's a specific reason to.

Reviewing your data sources

You'll also see your data sources listed on this page. These are the sources loaded into the Wordnerds platform. Like the other filters, the data sources you select here act as universal filters, if you choose a specific source, every other page will show data from that source only.

Understanding your project setup

Beyond filtering, this page gives you a useful overview of your project structure. On the left, you can see all the theme types set up in your Wordnerds project, upload themes, keyword themes, definition-led themes, and your frameworks. This helps you understand what categorisations are available for your analysis in Power BI.

Clicking through the theme categories lets you see which individual themes sit within each one, which is particularly handy for checking which metadata is associated with each data source.

On the right, there's a glossary of terms to help anyone who's less familiar with Wordnerds terminology get up to speed before exploring other pages.

Navigating forward

Once your filters are set, use the arrow button to move through to any other page in the report. Your selections will carry through automatically. You can always come back to this page to adjust your filters. Just click back to Data Structure, make your changes, and continue.

Key points

  • Purpose: Set up your analysis by selecting date ranges and filter options that carry through the entire report.
  • Starting point: Always begin here when opening your dashboard. We recommend keeping a standard set of filters for consistent analysis.
  • Universal filters: Data source, date range, volume type (author vs item), and blank response handling all apply across every page.
  • Project overview: View all themes, categories, and frameworks available in your project (visit Nerd Academy for more detail on themes and frameworks).
  • Glossary: Reference definitions of Wordnerds terminology for yourself or colleagues.
  • Navigation: Use the arrow button to move to other pages whilst keeping your filter selections active.
  • Volume vs author level analysis: Volumes will differ depending on your selection. If one person submitted ten responses, author view counts them once; item view counts them ten times. Sentiment and KPIs also change: item-level sentiment counts every piece of feedback equally, while author-level sentiment averages each person first, giving everyone equal influence.
  • Volume type guidance: Select item (verbatim level) for most analysis to ensure you're capturing every individual comment.
  • Including or excluding blanks:
    • Including blanks means the quantitative scores of people who left no text feedback are still considered.
    • Excluding blanks means those scores are not included in your analysis.
    • Blank responses are not assigned a sentiment score in Power BI. Note: this differs from the platform, where blank responses are given a neutral score by default.

Use case

Top tip! Use this page to establish a standard set of filters and try not to deviate unless there's a clear reason to. We recommend starting at item level and including blanks because all comments then hold equal weight in sentiment calculations, and the quantitative scores of people who didn't leave comments are still reflected in your analysis.

You can also use this page to confirm which metadata is associated with each data source. Select a data source, view the upload theme categories, then click into each one to see the individual upload themes (metadata) within it.

Zoe-1    ✍️ Article written by: Zoe, Customer Success Manager 

Still in need of some help? Give us an email on support@wordnerds.ai or reach out to your CSM directly.