Turn an IPA Reflection Into a Video Using NotebookLM (Full Transcript)

[00:00:00] Speaker 1: Imagine you are finished analyzing your data using interpretative phenomenological analysis and during that time you were able to produce your reflection based on the analysis that you did. So as you can see here I have a reflection generated based on the analysis for burnout among primary healthcare physicians. So as you can see here I have my reflection document here and I just want to do something very interesting. I want to transform this document into a short video and what I'm going to use is called Notebook LM. I was able to upload the document which is the reflection and created a video clicking on video overview and this is what I got.

[00:00:52] Speaker 2: The raw material of qualitative research is chaotic. You sit down with hours of recorded human experience, rambling thoughts, contradictions, and emotional pauses, and you are tasked with bringing order to it. You have to translate that tangled web of personal history into structured academic insight. To understand how researchers accomplish this we can look at a specific study analyzing five semi-structured interview transcripts. The subjects are primary healthcare.

[00:01:22] Speaker 1: It's just eight minutes. I've just played a little bit for you. I'm going to show you how I did this video. So the first thing that you have to do is that you have to have your reflection on a Word document. So I have my reflection here and then what you have to do is to go to Notebook LM. So we're going to go to Notebook LM. You just type Notebook LM and then you'll be able to go there. This is what you're going to see. You should have a Gmail account to be able to assess that so you can set everything up using your Gmail account. When you are ready you can go ahead and click on try Notebook LM and this is what you're going to see and what you're going to do is that you're going to click on a plus sign here and then you're going to upload the reflection or the document that you want to create a video based on that. So I click on upload here and then I look for the document. It will take a few seconds for it to completely upload. Now I have my document here about my reflections. Then what you have to do is to go to video overview. So this one will review the document and create five to ten minutes video about your reflection. So you click on that. You have three options here. You have the cinematic option which is the one that I like. It's more interactive. It's very rich in terms of the video and then you have an explainer and a brief. So I will choose the first option and here you have an option to also describe what kind of video that you expect right what should be the focus of the video. It's not required for you to put anything here but if you want you can let the system know about your intentions what you expect right your intent of the video that will give you more rich video that meets your expectation but you don't have to do that if you don't want to go into detail. It's not required but if you want to give a little bit of information to the system so that the video will be more customized then I want to also show you the kind of information that you want to give to the system so that it will really be helpful for you. That information can be grouped into five areas. You can provide a contest. You can provide assumptions, your perspectives, your intent, and also your expectations right. It doesn't have to be perfect. What you have to do is to provide some information that will help the system to know how the content of the video should look like. So the first one is a contest. Can you give the system a little bit of background information? Providing the system with background information will help the system to know what exactly you want so as to provide a video that meets your expectation. Also you can provide your assumptions right. What role do you want the system to play in this creating of the video? What are the constraints? So think about assumptions. It could be like claims or beliefs that should be considered when creating the video right. You also kind of provide your perspective. It's not required but what is the viewpoint right. In terms of the video will be like a narrative form so what point of view do you want the system to take as it's also generating the video. What role do you want it to take? Any information about perspective or the lens will be very helpful. Your intent. What is the goal of this video? What is the purpose of this video? That will help the system to provide you more information. And the last one is your expectation. What do you expect? What kind of format? What kind of tone? Is it academic tone or conversational tone? So these are the areas that you can think about. So think about CAPIE. You can call it CAPIE. Think of CAPIE whenever you want to provide detailed information to the system. But as I said it's not required for you to do that. But if you want to do it you can provide that information that will be also helpful. Let me show you an example of mine. It's not all that perfect. I generated this from Cloud AI. So this is the contest. So I provided a little context for the system to understand what the background of the document that I'm asking the system to use it to create a video. I provide assumptions right and then my perspective the intent in terms of what I want to accomplish and then my expectations. What I expect the system to do. You can even let an AI tool to generate this information for you. You can say that oh can you generate a prompt to create a video based on this component and the AI tool can generate something similar for you and then you can copy your pieces here right. If you don't put anything here still you'll be able to create a video right. So it's not like it's required. So when you are done you can click on create and then the system will create a video. It can take about 10 minutes to create so you have to be very patient. As we are waiting you can also click on infographic. It can create a visual representation about a document that you gave to the system. So in this one is your reflection. So you're going to create a visual representation right. So we will wait and then when it's ready I will show you the results right. So as we are waiting let's go back to the one that I've already created. So when we go here let me go back. Okay so this is the one. The first one I created here is I created it without putting any information about a contest. I just click on this one and click on this that one and I created. I didn't provide any information here and the second one I provided some information that will help the system to know what exactly I want. So you may ask why do you have to provide some information. So when we go to create what is the essence of providing some information here. So this is an opportunity for you to tell the system what exactly you want. You don't have to go through all the contests, the assumptions. Any information that will help the system to know what you want is good and then you click on create. As I said you don't have to do it. So the first one I did not provide any information. The second one I provided information. I think you saw a little bit of the first one. Let me show you the second one what I did.

[00:08:34] Speaker 2: When we hear physician burnout an image comes to mind. Endless charting, electronic health records and a mountain of administrative tasks crushing. Entering this study of five primary care doctors the researcher expected to hear exactly that. As an informed outsider the standard narrative felt like a foregone conclusion. But qualitative research requires a specific discipline called bracketing. You take your expectations and intentionally suspend them holding your own assumptions lightly to remain open to what the data actually says. This methodology, interpretive phenomenological analysis, demands a deeper look. The task is to move past a summary of complaints.

[00:09:19] Speaker 1: For me personally the second video which is the one that we just watched is more focus is in line with what I'm expecting the system to create for me. The first one was more general based on the data and it makes some assumptions about what I want. I like the second one because it's really provided information that I really need, emphasizing on the participants and also the specific cases and also comparing the experience to provide that information in my thought as I go through the data. I like the second one better than the first one. Let's go back and see what is going on. I think the system has been able to generate the info based on the data. Let's see. Okay, the data I analyzed is about Bernard Amon, primary health care physician. So this is my reflection. So it gives all the components about using IPA. The first one is about double hemaneutic. This means that you as a researcher want to give participants a chance to provide you how they make sense of the experience and then you also make sense of that information that they have provided to you. So it's a double understanding which is called double hemaneutic. And the second one is the individualized strategy where you analyze each data, develop each theme for each participant before you compare and contrast cases. And also another exploration is like focusing on the kinds of words that they use and the meaning or assumptions associated with the kinds of words that they use. And this also gives you how the findings look like, my thoughts about the findings. So this is really perfectly reflect what was in the reflection. Okay, so the video is ready. So let's click on it and see.

[00:11:18] Speaker 2: When I first stepped into the clinical life world to study physician burnout, I wasn't entering as a fellow doctor. I approached this as an educator and a qualitative methodologist, an informed outsider looking in. Like most people who read the literature, I arrived with a specific expectation. I assumed I would hear a standard narrative dominated by endless charting, administrative bloat, and crushing patient loads. But that assumption carries a glaring logical flaw. If clinical burnout is merely a symptom of systemic overload and bad scheduling, then adjusting a physician's hours should be an easy cure. Broad summaries of surface level complaints provide a catalog of grievances, but they miss the interpretive layer, the specific meaning these physicians attach to their exhaustion. To get past the surface, I used a specific qualitative tool, interpretive phenomenological analysis, or IPA. For this study, I applied its individualized strategy. The core rule of this strategy is strict isolation. You must complete the entire analytical cycle for one physician's transcript before you are allowed to look at a second case or attempt any comparisons. This diagram illustrates how the method forces the researcher into a practice called bracketing, actively holding back workload expectations. The filter catches the obvious complaints so that smaller, unrecognized anomalies can slip through for inspection. Working this way is demanding, but it is the required investigative tool to expose the exact details that broad, generalized surveys ignore. Take the first isolated case, a 37-year-old physician I'll call P1. When asked what caused her burnout, her response was surprisingly terse. She gave a short, clipped list of tasks, almost keeping the painful dimensions of the experience at arm's length. But her answer about how she coped was dense and rich. The heaviest analytical weight in her entire transcript rested on a single word she used when describing her desire for peace, again. This single word again carries significant weight. It implies that a state of peace existed before the exhaustion took over. If peace was the starting point, then burnout represents a disruption of that personal history. It transforms the experience into a sense of theft, the loss of an original baseline. Consequently, recovery requires restoration, an act of reclamation of the self that goes far beyond the benefits of mere rest. The most analytically demanding transcript belonged to P4. She was the youngest participant in the study, with only two years of experience, and she...

[00:13:50] Speaker 1: So you see how this video has brought my reflection to life, describing my thought process and how I was making sense of my data and the conclusions that I've made. This is so amazing.

[00:14:06] Speaker 2: ...experiencing burnout every single day. Yet paperwork was not the primary culprit. P4 was exhausting herself, performing the daily emotional labor of overcoming patient skepticism. As a young African-American female physician, she faced a credibility challenge written on her body before she ever spoke a word. There is a strong temptation in research to smooth over these outliers and absorb them into a broad category like workplace stress. I actively chose to preserve this distinct embodied divergence. The moments when a participant's lived reality strongly resists an emerging framework are the exact moments where the analysis yields the highest value. Then there was P3, a 70-year-old physician with 35 years of experience. He posed a completely different challenge because he barely experienced burnout at all. He created distance from the phenomenon by framing it as a hazard for younger physicians. He noted that without the right mindset, a doctor could easily fall. Fall is a highly specific verb. It moves the definition of burnout away from simple fatigue and recharacterizes it as a dangerous loss of professional footing, a structural collapse. His account proves that clinical burnout is not a fixed universal constant of the medical profession. It is a highly dynamic threat that hunts for specific vulnerabilities. Only after honoring these individual accounts in strict isolation did I lay the five theme tables side by side to look for cross-case connections. My initial expectation of systemic overload was replaced by a different universal finding. Every single physician in the study consistently attributed their burnout to external forces. Identifying that doctors blame external factors is a pattern. Making meaning out of that pattern requires understanding why they do it. This externalizing framework functions as an essential form of professional self-protection.

[00:16:11] Speaker 1: This universal pattern reveals a deeper architecture.

[00:16:14] Speaker 2: Clinical burnout functions as a profound struggle to maintain a coherent professional identity. The administrative and scheduling issues are the external pressures that make this internal identity work so exhausting. A timeline graph reveals a clear trajectory. Year two is defined by the acute vulnerability we saw with P4. Moving to mid-career, this evolves into complex restorative needs. Finally, at year 35, we find the near total absence of burnout, replaced by hard-won structural autonomy. This progression maps directly onto a single vital metric, a physician's gradual accumulation of professional agency over the lifespan of their career. Solving complex crises like burnout requires moving beyond systemic generalizations. We must commit to the rigorous individual interpretation of lived experience.

[00:17:10] Speaker 1: Okay, so what do you think about this video? If you have any questions, you can put them in the comment section. I'll be happy to address them for you. So after creating a video, one thing that you can also do is you can do a little bit of reflection, right? Did the video that you created meet your expectation? What are some of the limitations and also the strength in terms of the video representing the document that you gave to the system? You can also create a new one if you want to the same process, right? You can just click on video overview and then create a new one if you want, and maybe give the different instruction or you don't have to provide an instruction and see what the system will come up with, right? So this is what I have for you. If you have any questions, you can put in the comment section. I'll be happy to address them for you. Thank you so much for your time.