AI Tools for QDA

Information about genres of AI tools for qualitative analysis.

Information about other CAQDAS-packages

Genres of AI tools for qualitative analysis

General Purpose Chatbots
Gen AI Integrations into CAQDAS-packages
New Gen-AI Apps
Text Mining tools in CAQDAS-packages

Gen-AI capabilities harnessed for QDA analytic activities

Across the genres of AI-tools derived from the capabilities of LLMs are five "Gen-AI analytic capabilities". 

Generating

  • Using AI to generate ideas for aspects of the qualitative workflow.
  • Using AI tools for qualitative data collection.
  • Having AI generate 'silicone' or 'synthetic' data for use in qualitative research.

Converting

  • Speech-to-text capabilities used for automated transcription

Summarising

  • At different levels (e.g. whole documents, selected text segments, and already coded text segments).
  • This includes the function to explain terms i.e. if there is a technical term or theory in literature you’re unsure about you can have it give you an explanation - or if a participant in a transcript uses a colloquialism you’re not familiar with – but double check to make sure it’s not hallucinating.

Conversing

  • Chatting with qualitative data in sequential conversational chatbot interfaces, or in comparative grid displays.

Labelling 

  • Suggest codes from a selected text segment, or sub-codes for a code
  • AI-driven coding is enabled in three ways: 
    • (a) fully automated AI coding 
    • (b) based on human specified “intentions” , such as analytic questions, project context etc., and 
    • (c) ‘directed AI-coding’ which is based on a defined code – the AI then goes looking for text that fits, which in some cases it does surprisingly well. 
  • AI-driven theme  which is being offered by some of the new apps, but not currently in established CAQDAS packages.

These were outlined by Christina Silver and Steve Wright in their keynote presentation "The Good, the Bad and the Ugly of AI in Qualitative Analysis" at the Social Research Association Annual Conference in London on 6th May 2024. You can watch the recording and access the slides from here. 

 

General purpose Gen-AI Chatbots harnessed for QDA analytic activities

Some researchers are using general-purpose Chatbot-like tools such as ChatGPT, Claude, Perplexity etc. to facilitate aspects of the QDA workflow. 

Here we will be summarising these uses and linking to relevant resources. 

For now - check out these articles:

 

Gen-AI integrations into established CAQDAS-packages

This genre of Qual-AI is about established "traditional" CAQDAS-packages integrating the capabilities of Gen-AI into their existing suite of tools designed to facilitate qualitative (and mixed-methods) analysis. 

Currently these include:

ATLAS.ti - incorporating Gen-AI capabilities, from March 28th 2023

 

MAXQDA - incorporating Gen-AI capabilities, from April 26th 2023

 

QualCoder - open source qualitative software, incorporating Gen-AI capabilities, from 19th December 2023

 

NVivo - incorporating Gen-AI capabilities, from September 15th 2024

 

New Aps harnessing Gen-AI capabilities for QDA

Since late 2022 we have seen a plethora of new Aps come onto the market that are solely (or mainly) focused on harnessing the capabilities of LLMs and Generative-AI for qualitative analysis. The majority are coming out of the market research industry. But some are being developed by qualitative researchers with academic backgrounds and/or affiliations. 

There are way too many of these tools for us to usefully list here. To give context to that statement, we are receiving on average two or three cold-calls per week at the moment from new developers in this space, keen to show us their products. We literally do not have the capacity to familiarise with or review all of them. Indeed, there is little point in doing so, as likely many of them will not survive the current peak of the hype-cycle we're in. 

Therefore, what we're intending to do here is to distinguish between the types of new Aps and provide information on selected of them within those groups. 

Bear with us while we get to the point of having concrete examples to share with you.

Coming first will be (in alphabetical order):

  • AILYZE - Gen-AI qualitative data collection and analysis tool. Lead-development by James Goh, Boston, United States, who has background in development economics and data science at Massachusetts Institute if Technology (MIT) and The Wharton School (University of Pennsylvania).

 

  • CoLoop - Gen-AI qualitative data analysis tool. Lead-development by Jack Bowen, London, UK.
    • Our review of CoLoop is coming soon - watch this space!
    • Check out 10 minute overview from Jack on CoLoop at the Symposium on AI in Qualitative Analysis we organised with the SRA last year
    • Check out Episode 7 of our #CAQDASchat podcast when Jack and Christina chat about AI in qualitative research and the development of CoLoop.
  • Flowres - Gen-AI qualitative data collection and analysis tool. Lead-development by Jiten Madio (Founder and CEO), out of India
    • Our review of Flowres is coming soon - watch this space!
    • Check out the Flowres blog for more info.

 

  • MyRA - Gen-AI qualitative analysis tool. Lead-development by Alex Bish and Aliai Eusebi, London, UK
    • Our review of MyRa is coming soon - watch this space!
    • Check out the MyRa blog for more information.

 

  • QInsights - Gen-AI qualitative analysis tool. Lead-development by Susanne Friese, Rotterdam, the Netherlands

 

  • Reveal - an online AI-driven qualitative analysis platform that uses large language models (LLMs) to transcribe audio/video recordings of conversational data (e.g. interview and focus-group discussions) and to facilitate the their AI-driven analysis. A series of individual (participant-based) and study-level (comparative) analysis are provided in grid displays connected to underlying transcripts and (optionally) source audio-visual materials.
    • Reveal was founded in by Alok Jain and Niket Patel at Synthefai Inc. in Virginia (USA) and Gujarat (India).
    • See our review of Reveal (PDF).

 

Text Mining capabilities in CAQDAS packages

CAQDAS-packages have incorporated text-mining capabilities for many years, such as topic modelling, sentiment analysis and other forms of supervised and unsupervised machine learning. These capabilities pre-date Generative-AI capabilities that emerged from 2023. These pages focus on the explosion of Generative-AI into the field as a result of the availability of Large Language Models (LLMs) where the majority of debate is currently focused. 

For more on previous AI tools and their integration into CAQDAS-packages, see this post by Christina Silver, published on the QDAS blog 5th May 2023, for a brief overview.