Everyday AI Podcast – An AI and ChatGPT Podcast - EP 525: AI-Informed, Human-Led: Thoughtful AI Use in Qualitative Research

Episode Date: May 14, 2025

AI is shaking up qualitative research: speeding things up, cleaning messy transcripts, and even identifying hidden patterns.Sounds amazing, right? Buuuuuuuut there’s a catch. When does AI go from helpful assistant to heavy-handed editor, scrubbing out the human insights qualitative research was built on?In this episode of Everyday AI, we're tackling how to balance AI-powered qualitative research without losing touch with human nuance. Newsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageJoin the discussion: Have a question? Join the convo here.Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineupWebsite: YourEverydayAI.comEmail The Show: [e-mail] with Jordan on LinkedInTopics Covered in This Episode:Importance of Qualitative Research TrainingChallenges in AI-Assisted Qualitative ResearchAI's Role in Data TranscriptionHuman-Led Analysis in Qualitative ResearchAI's Impact on Data AccessibilityLimitations of AI in Qualitative AnalysisFuture of AI and Qualitative ResearchCritical Thinking in AI-Assisted ResearchTimestamps:00:00 PhD and Qualitative Research Training05:07 "Qualitative Research and Multiple Narratives"08:04 AI in Research: Balancing Automation09:57 AI-Assisted Data Analysis Caution16:23 AI's Impact on Qualitative Research19:19 Analyzing Qualitative Data Interpretation25:11 Balancing AI in Qualitative ResearchKeywords:Qualitative research, qualitative researcher, AI use in research, AI in qualitative research, AI-informed, human-led, thoughtful AI use, research impacts, health psychology, apprentice plumber analogy, training for PhD students, data analysis, transcribing interviews, transcription time, structured versus unstructured data, meaning making, words as data, interpretive work, subjective interpretation, data set analysis, line by line coding, thematic analysis, story telling in research, ensuring rigor in research, critical thinking in research, efficiency in research processes, literature gap identification, data visualization, AI as an intern, sharing research findings with lay audience, implicit meaning in data, biases in models, accessibility of academic work, academic pressure to publish, research dissemination, insights from research, qualitative research processes, interpretation in qualitative research.Send Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info) Start Here ▶️Not sure where to start when it comes to AI? Start with our Start Here Series. You can listen to the first drop -- Episode 691 -- or get free access to our Inner Cricle community and all episodes: StartHereSeries.com Also, here's a link to the entire series on a Spotify playlist. 

This is the Everyday AI Show, the Everyday Podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business, and everyday life. Meet Firefly AI Assistant, now live in Adobe Firefly, the All In One Creative AI Studio. Just describe what you want to create and the assistant handles the rest, orchestrating multi-step workflows across Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome. The assistant accelerates execution. Whether you realize it or not, research impacts our daily lives.

From healthcare improvements to product development to educational strategies and workplace policies, research shapes everything we do, experience our day-to-day activities. But when it comes to AI use in research, specifically qualitative, research. I think there's some aspects of this that we might be overlooking. So that's what we're going to be tackling today on everyday AI. What's going on, y'all? My name's Jordan Wilson. I'm the host of Everyday AI, and we are your daily live stream podcast and free daily newsletter, helping us all not just learn AI, but how we can leverage it to get ahead to grow our companies and our careers. If that's what you're trying to do, trying to be the smartest person in AI, your company or your

department, you're in the right spot. It starts here with this live stream. slash podcast, but it continues on our website. So if you haven't already, please go to your everyday AI.com. So there, sign up for the free daily newsletter. We recap the podcast each and every day with some exclusive insights in that newsletter, as well as keeping you up to date with everything else that's going on in the AI world. So it's a one stop shop for everything that you need. So make sure you go check this out. So if you are tuning it into the podcast, looking for the daily AI news. Technically, we got a pre-recorded one that we're debuting live. So all of that AI news will be in the newsletter. So make sure to go check that out. All right, enough. Chit-chat, y'all. I'm excited for

today's conversation. I hope you are as well. So, hey, for our live stream audience, can you welcome to our show? Dr. Claire Moran, qualitative research educator and facilitator. Claire, thank you so much for joining the Everyday AI show. Hi, Jordan. Thank you so much for having me. All right. I'm excited. I'm excited to talk this one. But first, before we get into it, tell us a little bit about your background as a qualitative research educator. We'll get into like what that actual like actually means, right? But tell us a little bit of your background. Sure. Yeah. So I'm actually Irish, first of all, as I should probably say, you can tell if I'm a accent. But I live in Australia and I did my PhD here in Queensland at the University of Queensland in health psychology. So my PhD is in

qualitative research. And over the last few years, what I've really now, in my field is there tends to be a bit of a lack of training for qualitative research, as well from PhD students. It's a kind of a funny situation. It's like if someone's going to be an apprentice plumber, but you said, okay, oh, if you go, learn on the job for four years, and then come back and I'll tell you if you're qualified or not. So there's not actually a lot of training for people who are doing PhDs. A lot of it is kind of trial and error, which can be really tricky and there's quite, you know, there can be kind of blind alleys and, you know, self-doubt and all that kind of stuff. So, you know,

a lot of what I do in my work is running training for people who are doing PhDs. A lot of people that come to my training are also seasoned academics. But basically it's a space for qualitative researchers to come together. And yeah, to do training, to learn from me, to learn from each other. So one of the most recent courses that I've started running is a AI and qualitative research training. Because again, as we know, the speed of AI is so incredibly fast. and institutions can be really slow to, you know, to move at the same pace. So we have this, you know, situation where we have all of these tools, all of this kind of

accelerated change. But researchers are often really unsure, well, you know, what does this actually look like for my research practice? You know, what's okay, what's not okay? So, you know, for me, part of doing training is, yes, the training. But there's a lot to be said as well for providing that space for people to come together and have those open conversations. And as I said, not just learn from me, but learn from each other.

And I think that's really powerful for people. And maybe let's start a little bit at the beginning. So maybe for those that aren't, you know, super in the know what it comes to how research works or, you know, what is a qualitative researcher? Let's start there. What the heck is a qualitative researcher? And how does it impact, you know, our daily lives? For sure.

So I think when people think of research, we all, we often, think of quantitative research, which is when data can be kind of reduced to numbers, okay, or we have a hypothesis and we're testing a hypothesis. And so when we're talking about that kind of quantitative research, you know, we're talking usually about, you know, definite answers or statistical significance. Okay. So, you know, you know, two plus two equals four. You cannot tell me two plus two equals five. Okay. So we have a definite and a clear answer. When we're talking about qualitative research, what we're talking about is meaning and meaning making. Okay. So what we're using as data, as qualitative researchers, is we're using words. And we can also be using things

like images, but primarily our main form of data is words. So what we're focusing on is things like processes and experiences and how, how things happen, rather than, then cause and effect and why. Okay. So within that kind of sense of meaning and meaning making, what we're also recognising is that as the researcher, we are telling one story of a number of stories that could be told about a data set. Now, when I use the word story, that doesn't mean we're not engaged in rigorous practice. Okay. So, you know, we're always striving to, you know, produce excellent research to do research in a really rigorous way, but nonetheless, being really clear on the fact that there is not one correct answer to a data set. If you imagine, for example,

if you had a, you know, if you had some Lego, if you had some bricks, and so we recognize that we can configure these bricks in a number of different ways. Okay, so, you know, which is correct? Is it this way or is it this way? So we recognize it's quite a stupid question. Because we can see that there's multiple ways that you can configure the bricks and all of those ways are actually correct. I love our podcast audience. Claire just actually had some, you know, jumbo Lego blocks. As weird as it is, I have a lot of experience with Jumbo Lego blocks. Yeah, another story for another day.

So, you know, one thing, and maybe help us understand this because I think that when people look at the quantitative side, they're like, oh, yeah, it makes sense for AI to play a huge role. in that, right? You know, when you have your numbers, your data, et cetera. But when it comes to qualitative, when it comes to making sense essentially of unstructured data, you know, where does AI play into this right now? So where are the, I guess the areas where it makes a lot of sense to use AI responsibly in qualitative research? Absolutely. So if we think about kind of, I suppose, the, you know, a research project. And so when, you know, any researchers embarking on a research project, they obviously understand there's a lot of work that needs to go into actually getting

the data in the first place. Then you have your data. So when you have your data and, you know, say, for example, if you're doing an interview or a focus group, you will have that audio data, which then needs to be transcribed. You know, like, so say before AI, people, I think, don't realize how time-consuming, for example, transcription is. So it actually takes about eight hours to transcribe one hour of spoken words, right? So it's incredibly time-consuming. And then we have, so obviously at the back end in terms of, you know, those kind of repetitive or organizational tasks, AI can be incredibly useful, okay?

Even in terms of, you know, if I'm designing a research project, I'm looking for, well, you know, what does, what, what literature am I missing? here, I've got some literature. Is there other literature in the field? Okay, so we can use, you know, different tools to actually, you know, to suggest academic papers, to summarize papers for us, also to identify potential gaps in the literature. Okay. So lots of ways that, but I think, again, what we've got to be careful there is that we are not outsourcing critical thinking. Okay. So if we think about doing research, critical thinking is obviously absolutely key. And what we don't want to be doing is dumbing down our process. What I always say to people when they

come to my training is you need to intimately get to know your data inside and out. Okay. And so, you know, again, when we're framing our project, that understanding of, you know, what is the literature, what is my research question? Obviously, nobody wants to be collecting data. And then later think, oh, God, the questions I asked are really flawed, or they weren't good research questions, or there biases that were actually, you know, really prominent in those questions. So again, we always need to use our critical thinking skills. And so, you know, when we're talking about those kinds of tasks, while we can obviously outsource tasks that are, you know, routine and that are, you know, helping us with efficiency, we need to be thinking about using AI like an intern, okay, or as like

an assistant. So AI is not the researcher. You are the researcher, but you can use a AI as an assistant. So as I said, you've got those kind of back-end tasks and then, you know, you've got your data. So, for example, then using AI to transcribe that data, incredibly helpful and incredibly efficient, the next step then is actually the analysis of the data. And that's where I think we need to be, you know, really careful. So, you know, one of the things I would do is if I was using AI as an assistant in data

analysis, I would analyze my data first. and then maybe ask AI to do an analysis on the data to see, you know, are there any fresh angles or is there anything I kind of missed, anything that's quite useful, anything that can kind of, you know, peak my thinking that I can go back into the data and look at it again.

I think there's a real danger of, because there's a lot of kind of, I suppose, discussion in the field and in different domains and also in grey literature around, you know, is it helpful, for example, to get AI to, you know, do the analysis and then you do the analysis. But obviously the problem there is, you're going to have, and the AI has really, really informed your thinking and how you interpret the data. So I think it's really important to actually sit and work through that analysis

yourself and then perhaps then bring an AI to look for those different angles. And then, you know, after we've done the analysis, then the next step then is actually writing up your research. and again, AI can be incredibly useful for helping, you know, helping structure our writing, helping with, you know, when our kind of, you know, when our writing is a bit messy or, you know, lots of different ways that we can have, you know, that writing support. So it might just even be refining what we've done, okay?

Or seeing, you know, there's ways that we might be able to improve our writing. we can also use AI as, you know, for visual tools. So again, you could put the data into AI and say, okay, here's some data, you know, put this into a table for me or a visual representation. So lots of incredibly helpful tasks. But I hope what I'm kind of illustrating is all of those tasks are the kinds of tasks that we could outsource to a smart assistant. We wouldn't actually be getting them to do the job.

Adobe just introduced an entirely new way to create, bringing the power and precision of its creative suite into one conversational experience. Meet Firefly AI Assistant, now live in the Adobe Firefly app, the all-in-one creative AI studio. Powered by Adobe's Creative Agent, Firefly AI Assistant lets you start with your vision, just describe what you want, and shape the outcome as it takes form with the Assistant. The assistant orchestrates multi-step workflows, drawing on 60-plus pro-grade tools across Adobe Creative Cloud apps, including Photoshop, Illustrator, Premiere, Lightroom Express, and more to help bring your ideas to life. You can also get started with creative skills, a growing library of pre-built workflows for common creative tasks, like batch editing photos, creating mood boards, portrait retouching, and creating social variations. Every step the assistant takes is visible, so you can refine,

redirect or take over at any time. You stay in the driver's seat as the creative director. Adobe Firefly AI assistant now in public beta. See it today at firefly.adobie.com. So a lot of things to unpack, but where my mind was going to, as you talked about this, is kind of drawing parallels. I mean, it's pretty unrelated, but I'm pulling out some things in this, right? So when I think of like AI-assisted drug discovery, right, like very different than qualitative research,

but, you know, you read all these stories now about what this could mean, right? Oh, it could, you know, mean instead of waiting, you know, 10 years and hundreds of millions of dollars for a new drug, you know, it could be a year and, you know, 10% of the cost. When it comes to research, you know, specifically qualitative research, how might proper and responsible AI, you know, change what's possible in our society, right? Does this just mean we're going to get, you know, better health and business outcomes? Does it just mean like we're going to get products that are much more personalized for people?

Like, what does this ultimately mean, like, if this is implemented in the right way? Like, how might, you know, AI really change qualitative research and what does that look like for us? Yeah. So I think, again, as I said, I think that, you know, if we think about the actual analysis is a central part of a research project. I think that really needs to be human-led. But I think that then the tasks that go before that,

the tasks that come after that, incredible in terms of speed and efficiency, identifying gaps, you know, synthesizing literature, identifying literature, identifying, you know, possible ways that this research could be extended. You know, what would be some kind of next step for this research,

Also, I think one of the problems that academia can often have is that it's often what academics produce is not necessarily accessible for a lay audience. And that's incredibly problematic because, you know, research is about people and it's about meaning. So it's really important that that meaning actually is able to be accessible to communities, you know, that recommendations can actually be implemented. So when we talk about things like accessibility and implementation, there can absolutely be barriers there. And academics like everyone else have got a million things to do and are often under pressure, you know, to publish this pressure, you know, for outputs is often teaching and lots of work, you know, that's required. And that can really limit people's capacity for, you know, looking at ways that we can actually make research more accessible. So I think in terms of accessibility, I think AI is incredible because, you know, looking at ways that we can actually make research more accessible. because, you know, we can use those tools to kind of tailor our, you know,

here, okay, this is my findings. I've done the analysis of, hey, how can I make this accessible for a lay audience? How can I make this accessible for this particular community? How could I make this accessible for a policy paper? So, you know, that kind of translation of research into practice. Thinking about, say, implementation, thinking about recommendations. So, again, what we can see is this central kind of core work of a researcher is the task of doing the actual research, which is literally sitting down and

analyzing the data. And within a qualitative paradigm, that is fundamentally interpretive work. So we recognize the interpretation, the subjectivity and the interpretation of the researcher are absolutely central to that. Okay. So getting AI to do the analysis is kind of at odds with that philosophical underpinning of qualitative research, but nonetheless, there are a lot of other tasks that are required when we do a research project. And as I said, even those steps of, you know, disseminating to wider audiences, that is not something that is necessarily always done, but it's something that is incredibly powerful and incredibly useful. And that is something that AI can massively assist with. So, you know, throughout this conversation, you've, you've, you've

kind of broken down the qualitative research process into a couple of different, you know, big bucket categories. And, you know, what I'm picking up on is you're talking about, you know, transcription or, you know, transcribing interviews, collecting data, analysis, identifying themes, writing the actual research paper, right? It seems like there's already some, you know, some no-brainer ways on safe and effective ways to implement AI and still, you know, uphold, you know, the qualitative research process. But what happens? as these models get better and better, right? So, you know, although they're newer, we do have these, you know, models that can kind of, you know, have this chain of thought, this thinking,

this reasoning, right, is probably not yet there to where we need for the level of research that we're talking about. But what happens as these models do get more robust and more capable and better with their analysis and identifying themes, right? So how should or how might the role and the kind of day-to-day of qualitative research, researchers, how might it change? For sure. So I think there's a couple of questions I'd like to mention there. So one is, okay, so just when we think about what AI does, we're talking fundamentally about pattern recognition, okay? And we're talking about things that happen frequently in the data. So what you're kind of noticing is those things that happen more frequently and identifying patterns, okay?

It's not actually making meaning. It's using an algorithm or a statistical frame to be able to kind of generate meaning and give that back to you. With qualitative research, frequency is not an indication of importance. Okay. So that's one thing that's really, really key. And secondly, often we find that there are biases. So if we think about obviously a large learning model, you know, that's been fed all of this data. And often the data that's been fed, there can be a lot of biases or things that,

maybe, for example, racial bias, sexist bias, homophobic bias. So we really run the risk of actually perpetuating biases in the data and noticing what is happening

frequently. So, like, you know, we're not so interested in. Again, it depends on what kind, I don't want to kind of get too technical here, but also kind of depends on what type of qualitative research and really what our research question is. So if I, let's say I was doing an evaluation of how people felt about a training, okay, really what I want is like a topic summary, I want a summary of people's responses, okay? So something like that, where really what we're

looking for is that kind of more surface level summary would be so useful to put into AI, okay, because what we can have is, okay, you know, this is the majority of people felt like this, other people felt like that. These are some kind of core things that people talked about. Okay. With kind of, you know, a qualitative kind of philosophy, what we're actually talking about is not so much what is talked about, but how things are talked about. Okay? So we're often looking at things like, you know, silences or contradictions and, you know, complexity in the data and that kind of depth in the data.

and it's not what comes across many, many times, but it's things that are, it's kind of how things are spoken about are kind of patterns across the data that don't actually speak to frequency, if that makes sense, okay? So that I think is where human interpretation is really key. And so, you know, the kind of work that I would do

with qualitative data is do a line by line coding. So what you're effectively doing is you're pulling a data set apart. So imagine like you've got a jumper or something, you've pulled the thread. You've unraveled the entire thing, right? You pull the data set apart. And then you're actually rebuilding it in a different form from the ground up. Okay. And that rebuilding enables you to start seeing different things in the data that weren't kind of obvious.

So I kind of think of it a little bit like pulling back a veil. Okay. So often like things that are kind of, you know, implicit in the data that are more that kind of latent level meaning. Can I just give you a quick example? Yeah, please. Okay, so I did an analysis on women's magazines. So what I was interested in laws, how is sex constructed in women's magazines and what are the implications for women in terms of sexual health? So we're talking about sexual health, the ability to, you know, negotiate condom use, for example.

Okay, so basically like, so sitting down with, so I had like 12 months of women's magazines and I identified all of the articles that were about sex and did kind of a line by line coding on all of those articles and then did the process that I just mentioned. So kind of rebuilding it in a different way, okay? And so when I did that analysis and I would present that analysis at conferences or whatever, you know, talk to people about it, people would say to me, now I understand why I hate women's magazines, but I've never been able to articulate. Okay. So it's like that thing of something that you kind of think, there's just something

here, but I'm just not sure what it is, right? Like, it's supposed to, they're supposed to make you feel good. They're supposed to be about, you know, freedom and independence, but they're just a bit yuck. And it's hard to put your finger on what that actually is. And so what my overarching theme of that paper was, it was something that was actually really, really small. So I ended up on one overarching theme which permeated the entire data set. Okay. That is not something that I believe that if I'd put that data into AI, it would have given me a very different analysis.

And so I think as a qualitative researcher, what you don't want is to tell people what they already know. What you don't want is to summarize data. You want to be able to pull back the veil, okay? you want to try and identify that implicit meaning in the data. And to be able to do that, again, our human interpretation is absolutely central. And as I said, that kind of unraveling and entirely pulling the data apart, rebuilding in a different way.

And so, you know, one of the things I mentioned to earlier, Jordan, was just that qualitative research is inherently time-consuming, it's messy. You have moments as a researcher where you just feel like, oh, my God, God, this feels like an absolute dog's breakfast. And I think as you're working through an analysis, if it doesn't feel like a dog's breakfast, you're actually doing something wrong. Because that kind of messiness and that kind of fog that can happen at many stages during an analysis are actually an inherently important part of doing qualitative research. Okay. So it is that kind of, you know, deep thinking and that kind of critical thinking and, you know, being able to sit in that

mess being okay with the ambiguity being okay with the mess and working through that enables us to come out the other side and and as research is one thing i always find when i do an analysis is i have moments where i'm like you know it's like when you're reading a novel and you have that plot twist and you're like oh my god i didn't see that coming and that's what doing good qualitative research should feel like okay you know you've done your interviews you've talked to the people you know you're familiar with everything the data says at a real surface level, when you actually sit down and you do that really deep thinking work of doing the analysis, you always have those aha moments where you're like, oh my God, I did not see that coming. And so that is the beauty of being a qualitative researcher and then being able

to share that with other people. And it really resonates. But again, it's that kind of behind the veil or that more implicit meaning in the data. And what we don't want to do is lose that. to efficiency, which is something different. Sure. All right. So we've covered a lot in today's conversation, Claire, but as we wrap up, what's your one most important takeaway when it comes to thoughtful AI use in qualitative research, whether that's advice for qualitative researchers or just

what you think is most important for, you know, our audience to know maybe as we are, you know, consuming the benefits or consuming what happens because of qualitative research, what's your one most important takeaway? Like, I think it is reflexive practice where, you know, we are understanding that subjectivity is really important as qualitative researchers keeping our own lens. So again, that human lens, that human interpretation, but really being able to use AI as that intern or that assistance,

that can really help us. with so many of those tasks. So I think, you know, that thing of sitting down and kind of mapping out the research process, what are all of the different tasks that, you know, will need to be done? What are the tasks that I can, you know, contribute outsource to AI

and have that kind of, you know, that collaboration between me and AI? What are the tasks that I need to really be in charge of as a researcher? And just being clear on what AI can do, what I can do, and how I can use AI to support me. But I think it's also,

So everything is very, very new. It's really, really important that we're having these conversations and that researchers are able to have access to ways to have the conversation, critical thinking and just critical awareness around and yeah, what the capabilities are, but also what some of those limitations are. I think is an important conversation to have that I think impacts way more of us than we actually might know. So, Claire, thank you so much for taking time out of your day.

share with us on the Everyday AI show. We really appreciate it. Thank you so much, Jordan. I really appreciate your time. Thanks, everyone for listening. All right. And as a reminder, y'all, we covered a lot. If you just need the takeaways or if you miss something that was said in there, don't worry. We've got it in our daily newsletter. So if you haven't already, please go to Your EverydayaI.com. Sign up for the free daily newsletter. If this was helpful, tell someone about it. Repost us. If you're listening on social media, if you are to listen on the podcast, please make sure to subscribe to the show. Leave us a rating. and also more importantly, join us tomorrow and every day for more Everyday AI. Thanks y'all.

Meet Firefly AI Assistant. Now live in Adobe Firefly, the Allman One Creative AI Studio. Just describe what you want to create in your own words and the assistant handles the rest, orchestrating multi-step workflows across Adobe Creative Cloud apps, including Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome while the assistant accelerates execution. Stand control with the ability to step in. in and refine at any time. See it today at firefly.adobie.com. And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going.

For a little more AI magic, visit Your EverydayAI.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.