Introducing Text Analytics
In the Text Analytics tab, there are many different tools to help you understand what employees are actually saying in their survey responses and interact with the results.
Using Machine Learning/Natural Language Processing, TINYpulse automatically analyzes all of the text responses (line by line) from your survey and categorizes them into different themes and sentiments.
Theme and Sentiment Axis Chart
The first thing you'll notice when you're viewing your text responses is the addition of an X-Axis chart.
Let's take a look and see how to interpret it.
The bigger the bubble → the more text responses there are regarding that topic.
- With so much text, it can be hard to quickly summarize what’s being talked about. Look for the largest bubbles to immediately get an idea of what topics are being talked about the most (from the text responses).
- The size of the bubbles correlate to the number of text responses tagged with that theme.
The further left → the more that topic is being spoken about in a negative light.
- In the example, Job Embeddedness is the most talked about topic in a negative way (from the text responses). That should immediately indicate that this is a topic to dig deeper into and address.
- The color of the bubble and its location on the X-axis correlates to the average sentiment of that bubble.
Filters include by question, date, sentiment, and theme.
- These filters, when applied, will update both the sentiment chart and the text responses sections (they are linked together).
Hover over the bubble to show the sentiment breakdown of all text responses tagged with that theme:
After reviewing the summary of the text responses by theme and sentiment, scroll down to read through the individual responses themselves.
You can sort this feedback by the filters you select at the top of the page or by clicking the bubbles. Click the three-dot menu next to the text feedback that you want to manage:
If you wish to categorize this feedback in a different way than what TINYpulse categorizes automatically for you, you can create your own custom tags and manually tag the feedback accordingly.
Send a message directly to the employee who wrote this response. The employee can respond back to you and you will be notified.
Add a note for other administrators (who have permission access to this feedback) to see. You can type “@” and type the name of a specific administrator if you want them to be notified of your note.
Setting the status of a piece of feedback will add it to the Wins Board, where you can organize and actionable feedback to make sure you’re staying on top of taking action.
Machine Learning/NLP will automatically assign a theme based on the computer’s interpretation, but it’s not always 100% accurate. If you find that a piece of feedback better fits a different category, go ahead and change it and the computer will learn from it to better assign the theme moving forward.
Machine Learning/NLP will automatically assign sentiment based on the computer’s interpretation, but it’s not always 100% accurate. If you find that a piece of feedback better fits a different sentiment, you can change it and the computer will learn from it to better assign the theme moving forward.