NVivo AI Cloud: Research Lifecycle

The research lifecycle in NVivo AI Cloud has two phases: you first create and frame a project, then run your data through the Analysis Pipeline, from importing sources, to organizing cases and variables, to generating initial topics and coding

Phase 1: Create a Project (Project Framing)

When you first open NVivo AI Cloud, you will define your research title and give some background about your research. The more information you give to AURA about your project, the more it can tailor its suggestions to your research.

A project is the top-level container for a piece of research in NVivo AI Cloud. Everything you work with, your sources, cases, variables, codes, memos, and the results of AI processing, belongs to exactly one project, and deleting a project removes everything inside it.

Creating a project is how you set up a new piece of research in NVivo AI Cloud. A project holds your framing, sources, codes, and memos in one place, so everything you do later, from importing documents to coding and writing memos, hangs off the project you create here. NVivo AI Cloud walks you through creation with a guided wizard, which collects your topic, research questions, methodology, team, and starting literature across five steps.

Project Framing steps:

1. On the Home page, select New project. If you have no projects yet, select Create your first project instead. The New Project wizard opens.
   

2. On the TOPIC step, enter a Title and a Description of your research problem, context, and goals, then click Continue.

   Note: NVivo AI Cloud detects the language of your text and uses it as the AI output language for the project. You can change this later from the project settings icon displayed next to the project title.

3. On the QUESTIONS step, review the AURA suggested questions and add any that fit, or type your own under Your Questions and click Add question -> Continue.  You need at least one question to continue.

4. On the METHOD step, by default, Initial Topics is selected. Optionally, you can list your Participants and click Continue.

5. On the TEAM step, enter the email addresses of collaborators you want to invite and click Continue, or skip the step.

   Note: No invitation will be sent at this point, because the project does not exist yet.

6. On the LITERATURE step, choose Find literature, then continue.
   Note: If you choose to skip, the project is created and NVivo AI Cloud opens it.

7. Review the abstracts and select the papers you want, then confirm your selection. The papers you confirmed are imported automatically as sources, with no separate upload.

   When you run the search, it lists results in order of relevance, e.g. 84% fit. You can view an article by clicking on the icon in the right corner, and it opens in the browser. If the article is open access, you can view the whole file to check that it is the right paper. 

8. Check the boxes next to the articles you want in your project and click Continue.
   The project is created and NVivo AI Cloud opens it.

Phase 2: Analysis Pipeline

Step 1: Import Source files and Summarize

Once your project is framed, you can import your research data. If you used Find Literature during project framing, you will see the articles you selected already listed in the Research Library section of the Import window, ready to be imported into your project. They will be marked as Found. You can also add to the list any literature that you have already downloaded onto your machine.

The Import window has two sections:

• Data: Where you will import files that will be the subject of your analysis. These can be interviews, focus groups, or policy documents (if that is what you will be analyzing). In the Data section, you can upload media files alongside documents: audio as MP3, M4A, WAV, OGG, WebM, and AAC, and video as MP4, WebM, MOV, AVI, MPEG, and MPG.

• Research Library: Where you import research materials that support your analysis. These can be articles from the literature, documentation that is relevant to the research, etc.

Steps to importing and summarizing your source files:

1. Click Add files, choose Data or Research Library option and pick files from your computer or drag them onto the Drag & drop files here area.

   Note: Each media file cannot exceed 4 GB in size and 5-hour duration.

   If you have uploaded audio or video files, do the following:

   1. Select the language to be transcribed and click Transcribe. Transcription turns a recording into text you can read, edit, and code.

   2. When the transcript is ready, click Needs review. The speakers in the transcript are identified as Speaker 1, Speaker 2, etc.

      Note: You can add the speaker's name by clicking on, for example, Speaker 1 and then selecting Create new speaker. Name the speaker press Enter (Windows) or Return (Mac) and then confirm that you want to add this new speaker name to all instances of Speaker 1. When you have finished adding speaker names you can check the transcript and correct any errors.

   3. Once you have finished, click Mark as reviewed on the upper right-hand corner.

2. Review all the files you selected for importation. Each file appears in the list with its name and an Uploaded state once it is ready. A file's category (Data or Research Library) can be changed after upload.  

3. Click Continue → AURA Reads.

4. Click View memo on a source to open its AURA-generated summary alongside the document. This is an important step as it is where you can check the AURA generated summary against the transcript and make your own edits to the summary.

   Wait for reading to finish. Sources become readable once the reading step completes.

5. You can now move on to organizing and coding them by clicking Continue → Organise. For more information, see Step 2: Organize - Cases (units of analysis) and Variables section below.

What happens when I import PDF, DOCX, TXT, and VTT files?

After import, AURA reads each document individually, going through the whole document. It then produces a summary memo. This includes a content overview, key topics covered in the interview, a relevance note explaining how the document connects to your research questions, notable quotes, and the document type.

You will be able to read the summary next to the document. There will be a split screen with the summary on one side and the document on the other side. You will be able to edit the summary memo as you read through the document.

Each summary is saved as a memo linked to its source and vice versa, so you can continue to review and edit. This step ensures AURA has genuinely comprehended your data before any analysis begins.

What happens when I import audio and video files?

Recordings are data too, and the Data section accepts them alongside documents: audio and video. The Research Library does not accept recordings. You will need to transcribe each media file to text so that you can read, edit, and code.

Once the transcription process is complete, the transcribed file needs to be reviewed. Reviewing is where you confirm the transcript reads correctly and update speaker and moderator details.

What are the languages supported by NVivo AI Cloud Transcription?

NVivo AI Cloud transcription supports 81 languages. Regional variants are included, for example 18 Arabic and 15 English dialects.

Afrikaans (af)	Albanian (sq)	Amharic (am)	Arabic (ar)	Armenian (hy)
Assamese (as)	Azerbaijani (az)	Basque (eu)	Bengali (bn)	Bhojpuri (bho)
Bosnian (bs)	Bulgarian (bg)	Burmese (my)	Cantonese (yue)	Catalan (ca)
Chinese, Mandarin (zh)	Croatian (hr)	Czech (cs)	Danish (da)	Dutch (nl)
English (en)	Estonian (et)	Filipino (fil)	Finnish (fi)	French (fr)
Galician (gl)	Georgian (ka)	German (de)	Greek (el)	Gujarati (gu)
Hebrew (he)	Hindi (hi)	Hungarian (hu)	Icelandic (is)	Indonesian (id)
Irish (ga)	Italian (it)	Japanese (ja)	Javanese (jv)	Kannada (kn)
Kazakh (kk)	Khmer (km)	Korean (ko)	Lao (lo)	Latvian (lv)
Lithuanian (lt)	Macedonian (mk)	Malay (ms)	Malayalam (ml)	Maltese (mt)
Marathi (mr)	Mongolian (mn)	Nepali (ne)	Norwegian, Bokmål (nb)	Odia / Oriya (or)
Pashto (ps)	Persian / Farsi (fa)	Polish (pl)	Portuguese (pt)	Punjabi (pa)
Romanian (ro)	Russian (ru)	Serbian (sr)	Sinhala (si)	Slovak (sk)
Slovenian (sl)	Somali (so)	Spanish (es)	Swahili (sw)	Swedish (sv)
Tamil (ta)	Telugu (te)	Thai (th)	Turkish (tr)	Ukrainian (uk)
Urdu (ur)	Uzbek (uz)	Vietnamese (vi)	Welsh (cy)	Wu Chinese / Shanghainese (wuu)
Zulu (zu)

 

Step 2: Organize - Cases (units of analysis) and Variables

The organise step is your chance to give the corpus a case-and-variable structure before AURA reads it for topics. Organizing data early keeps the qualitative structure researcher-driven and improves the downstream analysis. The cases and variables can be created either during the AI Analysis pipeline or after you have been through the pipeline.

Steps to create cases and variables:

1. In the Organise screen, you can choose one of the two options:

   • The One case per document option creates a single case for each source, named after that source. It is the right choice when a variable's value applies to a whole document, e.g. publication year, author, etc.

   • You can use Analyze with AURA to identify and create cases for speakers, and if the variable information is included in the transcript, it will also create the variables and assign the values to the cases. If you have focus groups, interviews, or transcripts of meetings, use this way so each speaker is created as a case.

     Note: The CASES & VARIABLES grid sits at the bottom of the Organize step and is the manual path through it. Use it to create cases and variables directly and to fill in values yourself, whether or not you used one of the above options first. The matrix table is the case-and-variable editor embedded here so you can organize without leaving the pipeline.

2. Click Continue → Initial Topics.

 

Step 3: Initial Topics and Coding

The Initial Topics step turns your corpus into a starting code frame. AURA identifies recurring topics across your documents and finds excerpts to code to them. You can choose how topics should be sourced and start with your analysis.

Steps to topic creation and coding:

1. When you advance out of the Organize step, you can choose from the following options:

   • AI initial topics - You can ask AURA to identify initial topics, and you will be able to confirm or reject them after reviewing a sample of the text for each topic. AURA will then code the confirmed topics.

   • I'll provide my own - You can create your own broad topic areas with definitions, and AURA will code your data to them.

   • Both - You can use a combination of the above options.

2. Click Start.

3. Review the topics.

4. Confirm or reject topics as required.

5. Click Done to complete the pipeline.

NVivo AI Cloud marks the pipeline as complete, and your full project workspace is displayed.