(This feature is only available in NVivo installations with coding enhancements enabled.)

Automated insights

The term 'automated insights' in NVivo refers to the autocoding feature that automatically detects and codes themes or sentiment.

Use automated insights when you have limited time or large sets of data to develop preliminary results at the start of a project, or validate your thinking as you work. The automated insights functionality enables you to do approximate 'broad-brush' coding of large volumes of text quickly, which you can then review and refine.

Automated insights lets you work through large sets of data more efficiently—so you can get quick results to see what is happening in your content. You can focus less on the process of coding and more on the analysis of the results.

Automated insights is particularly useful if you have a very short time to form a broad conclusion. For example, a research consultant could quickly process public submissions to report findings in a timely manner during public consultation and program evaluation.

Be aware that there is a trade-off with automated insights—if time is paramount, you will get quick results but the results could be less accurate. If accuracy is paramount, you should code manually.

You do not need to have an understanding of data mining, text analytics or statistical methods and their terminology. There is no need to build up categories or criteria prior to running the analysis. The process involves selecting a set of files and using the Autocode Wizard to produce results. In a few simple steps, the Wizard makes it easy for you to gain an understanding of your data.

You can choose to autocode your files and codes for themes or sentiment.

Use automated insights as a discovery tool

Aimed at commercial and government researchers, automated insights provides a first-step in getting started in an NVivo project.

Bring in your files, then autocode themes or sentiment as a first pass over the data. This will assign a broad categorization to the data and provide some initial building blocks for you to work with. This results may be sufficient to draw your attention to areas of interest. By autocoding themes you should be able to quickly discover patterns or trends in your files.

Once you have some initial coding you can use the results to:

  • Code-on to develop ideas by coding content to other codes.
  • Identify potential areas for deeper analysis.
  • Frame questions for further data collection.

If you have a dataset of survey results, you can use the Survey Import wizard which automatically codes survey responses to codes, then detects themes and sentiment in the responses.

Process for auto coding using the Survey Import wizard.

Use automated insights as a verification tool

Once you have done your own coding you could run automated insights to validate your thoughts. For example, automated insights may reveal elements within the data that you may have missed or overlooked. It can also act as 'another set of eyes' on your files.

Since automated insights is computer-driven, the results are not influenced by your own bias.

Limitations of automated insights

Automated insights is not intended to replace the process of reading your files. You still need to be familiar with your content to achieve meaningful research results. This feature is unlikely to be helpful when dealing with secondary data, for example, literature reviews.

Automated insights does not perform analysis on the results. You will need to perform further steps like running a matrix coding query to draw meaning from the results.

You cannot predetermine the themes that are searched for, and the 'stop word' list is not used when autocoding themes and sentiment.

Running automated insights doesn't limit your coding and you're not locked in to the results. The codes are created in a designated folder, separate from your own codes. You can move and merge codes that result from autocoding themes, and you can easily change sentiment coding if you want to.

NOTE  There are some limitations when processing very large documents (or very large sets of documents) on the 32-bit version of NVivo. In this instance, install the 64-bit version of NVivo to use automated insights. You can see which version of software you are using on the About dialog box. (File > Help > About NVivo).

Privacy of your data

NVivo uses a proprietary tool to process the data—and the solution is fully integrated with NVivo. All data is held and processed within NVivo, so the security and privacy of your data remain within your control.

Projects in other languages

The automated insights functionality can only process one language at a time. This language is determined by the text content language setting of your project.

You should select files of the same language to analyze. Make sure your project text content language is set to the language of the sources you are analyzing.

The following languages are supported:

  • Chinese (Simplified)
  • English
  • French
  • German
  • Japanese
  • Portuguese
  • Spanish

If your project text content language is set to "Other", you won't be able to use automated insights.

The automated insights process may produce different results in different languages. If you have a translated version of the same file, analyzing the French version in French and the English version in English may produce different results.