Displayr

Example: a tracking survey where the user must translate the responses for the new wave of data. Re-running the AI Text Translation variable with new responses can change previously translated responses and impact Text Classification. Add an option for the user to be able to run the AI translations only on new responses, while keeping all previously coded responses intact.

"DisplayR does not allow these AI summaries / text to be edited in PPT – they are exported as images / pictures. The workaround for this is to create a text box and paste in the text here which on export can be edited. However, this is reducing the automation of the DisplayR workflow. Next month we update with a new data file, we need to manually copy and paste the AI generated text from below and past into the text box, as per the below. Also, the AI summary text does not allow other font sizes and fonts to be used – so we cannot use them in exports as pictures as they don’t allow our custom fonts."

The auto-updating interpretation outputs do not allow for any customization of the text. This is understandable, given that it auto-updates. However, it would be great if there was some way to pull out the bits from this that I want to show to my customers, i.e., that I would be able to reference parts of the outcome as required in a separate text field or output. Or, perhaps, specify exactly how I want the outputs from the function to read/look.

I would like a built-in tool that can detect if an open-ended comment was generated by AI. It would be helpful to have a likelihood score variable to indicate the probability of AI generation. This feature would enhance the ability to identify AI-generated data more accurately.

In the data preparation agent, there is an option to "Create a better name for the variable sets, including a number if it exists (e.g., How old are you? -> Q3. Age)." It should be possible to select both things separately, i.e., an option for: "Add variable name(s) to the beginning of the label where these can be determined" and "Create a better name for...." (as is but without the number part). As a user, I may well want one, but not the other. See also the feature request for "Show variable names in front of label" which works on a variable level, but not for grouped variables. That is, it would be better to have a labelling option that can do this intelligently.

It would be beneficial if the AI tool could read uploaded PDFs alongside survey data to inform its outputs. This integration would enhance the tool's analytical capabilities by incorporating more comprehensive data sources.

When combining variables Q8a_1, Q8a_2, Q8a_3... it names the new variable 'Q8a Factors that impact XYZ', rather than just 'Factors that impact XZY'. Currently, it doesn't include the question number in the new variabl,e which makes it difficult to find.

Please automatically put the results of the Research Agent into its own folder.

When you right-click > Interpret Data, you have to re-enter the Background section of the prompt. It would be good if the overall goals of the research can be stored somewhere which prevents having to manually re-enter this each time you run this feature.

The AI Token Usage report tracks the number of tokens used, but there's no means to tie this back to actual activity/actions in a document. I.e., it needs to have the document ID or Name, and also who was working in the document at the time / took the action that cost Tokens.