Writing Up Your Results

1:11Well, welcome to this deep dive. Today, we're taking you into the mechanics of that shared reality. We are, we're pulling apart chapter 15, writing up your results from the textbook research methods, from theory to practice. Yes, a very crucial chapter. Because our goal here is to really decode how researchers take complex, messy data and transform it into an evidence -based narrative that like actually impacts the field.

1:38Because writing up your results, it's not just about, you know, formatting your citations correctly. It's the ultimate stress test of the research itself. Exactly. It really is a stress test. The writing process kind of forces you to clarify your own logic. Like if you can't explain your methodology seamlessly, it usually means there's a flaw in the method. Exactly. There's a crack in the methodology itself. I, uh, I kind of like to think of this process, like handing down a really highly technical family recipe, you know, you can't just give someone a photo of a beautiful cake and say, like, figure it out.

2:10Right. The photo doesn't help you bake it. Yeah. And you can't just write bake until it looks done. You have to specify that you need, uh, exactly 200 grams of a very specific brand of flour and the oven must be exactly 350 degrees for 42 minutes. Yeah. Because if you leave room for interpretation, the next person's cake is going to completely collapse. Exactly. It's going to flop.

2:32That is a really highly accurate way to look at it, especially when we talk about replication. But before we get into the exact measurements of the paper, we kind of need to calibrate the overarching mindset. Okay. Like who is actually trying to bake this cake? For most researchers, the target audience is a sub -discipline journal. Meaning what? Exactly. Meaning you are writing for seasoned experts. They read hundreds of these papers.

2:55They have staggering cognitive loads and they really just want the conceptual core of your work. Like immediately makes sense. So the ticks lays out this guiding philosophy. They call it the three C's. Your writing must be clear, concise and compelling. Okay. I actually want to pause on that third compelling. I can imagine a lot of scientists, uh, balking at that word. Oh, they do all the time.

3:18Right. Because isn't the whole point of science to be completely objective and detached? Like if my data is rigorous, why is it my job to entertain the reader? Well, that's the thing. Being compelling in science has absolutely nothing to do with entertainment. Oh, really? Yeah. It has everything to do with momentum. You know, readers have very limited cognitive endurance. If you force them to wade through these dense convoluted meandering paragraph.

3:44You're just going to be all on the paper. Exactly. They will abandon it before they ever reach your brilliant conclusion. So a compelling scientific narrative is just one that effortlessly guides the reader from the broader context right down to the specific variables. It makes it painfully obvious why this specific research matters right now. So it's really about reducing friction for the reader. Yes, precisely. Which actually brings up a massive stylistic shift that the text addresses because there's always been this stereotype that to sound, you know, scientific, you have to completely strip yourself out of the text and use the passive voice.

4:18Oh, the dreaded robotic science voice. Yeah. Like the participants were observed or the compounds were mixed. It sounds like the lab equipment just spontaneously decided to do the experiment on its own. Right. It's this very persistent misconception that the passive voice somehow equates to objectivity. But the APA, the American Psychological Association, they actually actively fight against this. Wait, really? The APA fights against it? Yeah. They explicitly encourage using the active voice and using personal pronouns.

4:49But doesn't writing we observe the participants sound kind of inherently biased? Like the researchers injecting their own ego into the data or something? I mean, it might feel that way at first, but actually it's way more transparent because the researchers did observe the participants. Pretending some invisible hand conducted the study is just a linguistic illusion. That's a good point. Using the active voice forces researchers to take direct ownership of their methodological choices.

5:15And plus, from a purely mechanical standpoint, it just prevents your sentences from twisting into these awkward, unreadable knots. That transparency is crucial, especially when we talk about the difference between a narrative and a story, because the textbook draws a really hard line here. It is a vital distinction. In casual conversation, we use those words interchangeably all the time. But in scientific writing, they are just worlds apart.

5:40Right. A narrative flows logically from structured, verified evidence. A story, on the other hand, implies a beginning, a middle, and an end that has been neatly crafted, you know, often by smoothing over messy details. Like fiction, basically. Yeah. Yeah. The book references Dietrich Staple, which is this notorious case of data fraud in psychology. He fabricated data entirely because real world data is rarely neat. He wanted to tell these perfect, compelling stories.

6:08Wow. So he prioritized the plot over the actual truth. Exactly. So, okay, our goal is an evidence -based narrative. But what physical shape does that narrative actually take? Because the textbook uses a very specific structural metaphor to guide the writer. Right. Yes. The hourglass organization. If you just visualize the physical shape of an hourglass, it perfectly maps onto the conceptual flow of a research manuscript. Okay, let's trace that shape for everyone listening.

6:35The top of the hourglass is wide. This represents the introduction, right? Where you start broad. Correct. You introduce the general concepts and situate your specific question within the wider landscape of past research. But the text warns against zooming out too far. Like if you're researching modern cognitive behavioral therapy, you don't need to start your paper by citing Wilhelm Wundt, opening the first psychology lab in 1879. Right.

7:01Please don't do that. You provide relevant context, not a comprehensive history of the entire discipline. Right. Then as you move down the hourglass, the sides taper inward into the narrow neck. And this represents your methods and results sections. This is the most specific, cut and If you survive that narrow neck, the hourglass flares back out at the bottom. And this is your discussion section. You take those highly specific, isolated results from the neck and you broaden back out to discuss what they actually mean for the real world.

7:32Spot on. And throughout this entire structure, the guiding principle is finding what the book calls the right level of detail. You know, you need to provide enough precision so an independent lab could replicate your exact parameters, but you have to just ruthlessly cut the fluff. Right. The book gives a really great example of a terrible sentence. It goes, Professor Benjamin Franklin of the Department of Psychology at the University of Philadelphia in 1776 wrote a paper.

7:59It's exhausting just listening to it. Precisely. All of that is useless friction. You should just write Franklin 1776 found and let the references section handle the rest. You know, the hourglass metaphor, it really reminds me of zooming in and out on a digital map. Oh, how so? Well, you start your paper looking at the view of a whole country. That's the broad research context. Then you zoom all the way down to a single street view.

8:22That's the narrow neck where you detail the exact methods and statistical results. Right. Right. And finally, in the discussion, you zoom back out to the city level to show how, like, closing that one street affects the traffic of the entire metropolitan area. I really like that analogy. And if you zoom out too far in your discussion, say, claiming that closing one street will fix the global economy, you've lost the plot entirely.

8:47Yeah, that's way too broad. You have to broaden out, but stay within the absolute limits of your data. Okay. Let's look at how this plays out in practice, because the textbook anchors this entire process to a real world example. Right. Stella Christie and Deidre Jettner's 2010 award winning study on how children learn relational abstractions. Yes. So starting at the very top of the document, the title page and authorship.

9:10It's really crucial to understand that an authorship line is not just a polite list of who was in the room. Order matters immensely. Okay. The primary author, the first name listed, is the person who conceptualized the core research question and drove the main intellectual labor. Which actually brings up a really common friction point that a lot of students run into. If, say, an undergraduate student is working in a massive university lab and the principal investigator secured all the grant funding and they own all the equipment, how does the student ever get to be the primary author?

9:44That's a great question. The text uses a quote from Stella Christie herself to settle this. She explicitly states that ownership of the idea and the primary intellectual design dictates the first author. Interesting. Yeah. Doing the mechanical labor, like running the software, scheduling participants, manually entering data that does not warrant primary authorship, it might not even warrant authorship at all. Perhaps just an acknowledgement. Wow. So authorship is really about the intellectual architecture, not just laying the bricks.

10:15Exactly. And this is why researchers have to have incredibly frank and sometimes uncomfortable conversations about authorship before the first piece of data is ever even collected. Right. Get that sorted early. So directly below that author line is the abstract. And this is, what, a 150 to 250 word summary of the entire paper. That's it. Which feels incredibly restrictive. Why is the word count so violently strict? Well, two reasons.

10:39First, cognitive load for the reader, like we discussed earlier. Second, digital indexing systems. Databases use these abstracts to categorize and surface research. If it's too long, it literally breaks the system. You have to distill your purpose, participants, methods, results, and implications into less than a single page. That's so tough. It is. But Christie and Jettner, they managed to capture their entire complex study in exactly 158 words.

11:06And the ultimate pro tip from the text, always write the abstract last. Oh, really? Why? Because you cannot accurately summarize a narrative you haven't actually finished structuring yet. Oh, yeah. That makes total sense. Yeah. So the abstract took some, and now we have to guide them into the weeds. We move into the introduction, setting up the theoretical pins, and then we drop right into the narrowest part of the hourglass, the methods.

11:30Yes. The methods section is essentially the blueprint of the study, and it demands relentless precision starting with the participants. You don't just say like, we tested some kids. Right. You have to document the exact demographics. Christie's paper notes, the participants were three year olds, but you also have to document attrition. Attrition meaning the dropout rate. But wait, why is documenting the people who didn't finish the study just as important as the ones who did?

11:56Because silently dropping data skews reality. Think about it. If 30 people start a medication trial and 15 dropout because of severe nausea, but you only report the data for the 15 who felt great. Oh, wow. You have fundamentally misrepresented the safety of the drug. Now in Christie's field, infant and toddler cognitive development attrition is just a reality of the demographic. Right. Toddlers are unpredictable. Exactly. Toddlers get hungry.

12:24They throw tantrums. They fall asleep. You have to state exactly how many dropped out and why so the reader knows your final sample is truly representative. It's kind of like checking the structural integrity of the experiment. If you don't document who was in the room, nobody knows if your findings apply outside of that room. Perfectly said. Okay. So next, the methods section requires you to detail the materials and procedures.

12:46You have to explain exactly what the participants experienced. It's like a casting call and a script for a movie. Yes. The text actually reproduces Christie's visual stimuli. She was testing if children could recognize relationships between objects. So she showed them standards, for example, an image of a black cat sitting above a white cat. Okay. Then she gave them two options to match it. One was an exact object match and the other was a relational match like a black bird sitting above a white bird.

13:14Got it. Every single detail of these images, their size, their color, and the exact script the researchers read to the toddlers, it all has to be explicitly documented. Okay. So the blueprint is drawn. We know the participants. We know the exact materials. Now we face what is arguably the hardest translation job in the entire scientific process, the results section. Oh yeah. It is notoriously difficult because you are tasked with turning dry, highly complex statistical mathematics into readable prose.

13:42And the textbook recommends relying heavily on data visualization here to do the heavy lifting, right? It does. It highlights Christie's figure 15 .10. It's this bar graph showing the proportion of relational matches made by the three -year -olds versus the four -year -olds. Yeah. But reading these graphs, it kind of requires knowing the shorthand. Yes. There is definitely a visual vocabulary. For instance, you'll see these error bars extending out from the top of the bar graphs.

14:06What do those mean? Those bars represent the variance or like how spread out the data is around the average. If the error bar is massive, it means your participants were acting wildly different from one another. Oh, okay. You also look for asterisks. In statistical shorthand, an asterisk usually indicates a statistically significant difference between the groups, typically at a p -value of less than 0 .05. Conversely, if you see n's, that stands for non -significant.

14:32Okay. And then we get to the actual text output, the really dense statistical sentences. The book gives an example that looks like this. F statistic of 1 in 52 equals 23 .83. P is less than 0 .001 and eta squared equals 0 .30. Now, if you aren't a statistician, your eyes just completely glaze over reading that. We know the piss poo being less than 0 .001 tells us the result is statistically significant and like highly unlikely to be a fluke.

15:01But what is that F statistic actually doing? So the F statistic is fundamentally a ratio of signal to noise. Signal to noise. Yeah. It measures the variance between your groups, say, did the four -year -olds perform drastically differently than the three -year -olds and divides it by the variance within the groups, meaning how differently did the three -year -olds act from each other? Ah, I see. If the difference between the age groups is huge and the difference within the individual age groups is tiny, you get a large F number.

15:29It means your signal is overpowering the noise. And the 1 in 52? The 1 in 52 in the parentheses are just the degrees of freedom, which essentially tells the reader the size and structure of your data set. Okay. Wow, that makes it so much more approachable. It's really just a signal to noise ratio. But the text emphasizes that the F and the P value aren't enough on their own.

15:48You also have that last symbol, the eta squared, the effect size. Yes. Effect size is arguably the most important number in that whole string. Really? Why? Because a P value just tells you that a mathematical difference exists. But the effect size, the eta squared, tells you the actual magnitude of that difference. It tells you if the difference is large enough to actually care about in the real world.

16:12Oh. Yeah, you can have a statistically significant finding that is so microscopically small in its effect that it has absolutely no practical application. Here's a reality check, though. Yeah. If a researcher spends like six months running 50 different statistical analyses on a massive data set, there is a huge temptation to include all 50 of those tests in the results section. Yeah. You want to show your work.

16:37You want the reader to know how hard you worked. It's a very common urge, but it's a massive mistake. Yeah. The text is very clear on this. You should only report the statistics that directly address your main hypotheses. Dumping every single test you ran into the manuscript just creates massive cognitive overload for the reader, and it completely muddies the narrative. So you just leave the rest out.

16:57Well, you must report the analyses related to your hypothesis. Even if the data completely contradicts your predictions, you can never hide inconvenient data, but you have to stay focused. Okay. So we have the signal and we've proven the signal matters. Now the hourglass flares back out. We enter the discussion section where we interpret what these precise results actually mean for the broader world. But this is also where researchers are, I guess, most prone to slipping up.

17:24The textbook flags three major pitfalls here. Yes. So the discussion is where you summarize your findings, you acknowledge your limitations, and you address alternative explanations. Christy, for example, had to address a competing theory called cross situational learning, and she used her specific data to explain why her results ruled that theory out. Right. But when interpreting data, pitfall number one is by far the most dangerous. The word prove.

17:50Oh, I see this constantly in the news. This study proves that. You should practically strike that word from your vocabulary. A single scientific study cannot prove anything definitively. You should use words like indicate or suggest. I kind of want to push back on that though, because to the general public, avoiding prove sounds like a lack of confidence. If you have the data, the F statistic is massive.

18:11The effect size is huge. Why are scientists so terrified of just claiming victory? Because science is an iterative mechanism. It's not a final decree. A study provides evidence, sometimes overwhelmingly strong evidence, sure, but you must ethically leave the door open for a future more sophisticated methodology to find a more nuanced explanation. I see. The exact moment you say prove, you are essentially claiming the scientific process on this topic is finished.

18:39And it never is. That makes a lot of sense. So pitfall number two is anthropomorphizing, giving human actions to inanimate concepts, like don't write the research found. Data does not wander around the lab discovering things. You have to write the researchers found. It comes back to that active voice humans are responsible for the findings. Exactly. And pitfall three deals with significant figures. This is where your grammatical precision has to match your instrumental precision.

19:04What does that mean in practice? Well, if your measurement tool only outputs whole numbers, say a questionnaire where participants can only score an 85 or 96, and your statistical software spits out an average of 93 .7521, you cannot report all those decimal places. Because it creates this illusion of hyper accuracy that didn't actually exist in the real room. Exactly. If your tool wasn't that precise, you're right and can't pretend it was, just report the mean is 93 .7 and move on.

19:32Right. Every single word and number is a representation of reality. If the representation is flawed, the shared reality breaks down. And what happens when that shared reality breaks down? I mean, why is there such a relentless, almost obsessive focus on avoiding the word prove and not overstating your claims? Because once a research paper leaves the quiet sanctuary of an academic journal, it enters the public domain and the media can easily misinterpret it with massive real world consequences.

20:00Yes. The text highlights a fascinating case study about this regarding a working paper by David Filio at Northwestern University. It perfectly illustrates why this hyper precision really matters. Right. So Filio's study was looking at teaching effectiveness. His data indicated that students actually performed better in subsequent courses when their initial class was taught by non -tenure line lecturers. Now, within the incredibly specific context of his study at Northwestern, these lecturers were full -time, long -term faculty members.

20:29They were hired specifically for their elite teaching abilities. They just weren't on traditional tenure track. But when major media outlets picked up the working paper publications like the Atlantic, the Wall Street Journal, inside higher ed, they conflated his terminology. They incorrectly reported that the study showed part -time adjuncts were better teachers, which is a completely different employment category. Adjuncts are often part -time, highly precarious workers, cobbling together classes across different campuses.

20:56By confusing full -time lecturers with part -time adjuncts, the media inadvertently sparked this massive, politically explosive debate about university hiring practices, the exploitation of adjuncts, and faculty salaries. And, you know, just to be completely clear to you listening, we aren't weighing in on the realities of the adjunct system today. No, not at all. We are looking strictly at how the data itself was weaponized. Filio was stunned because his paper never even analyzed part -time adjuncts.

21:26It is a terrifying lesson for a researcher. Filio concluded researchers have a sacred duty not to draw outlandish inferences and to define their parameters so tightly that misinterpretation is nearly impossible. Because even when you are careful, the press or the public might distort your findings to fit a pre -existing narrative. If your writing has even a millimeter of ambiguity, you leave the door open for your data to be hijacked?

21:52Sacred duty. That really elevates the stakes, doesn't it? Writing up your results isn't just a chore. It isn't just some administrative hurdle to secure a grant or a publication. It is a fundamental public responsibility to protect the integrity of the knowledge you've uncovered. It is the ultimate defense mechanism for the truth. You craft that clear, concise, compelling hourglass narrative. You ruthlessly define your terms to avoid media conflation.

22:16You present your signal to noise ratio clearly, and you let the evidence speak without over claiming. It is precision executed at every single layer of the document. From the intellectual ownership on the title page, down into the 150 -word abstract, plunging into the narrow neck of demographic attrition and effect sizes, and finally, flaring back out into a discussion that is, frankly, confident enough to acknowledge its own limitations.

22:41And when you execute that structure perfectly, you guarantee that your discovery doesn't end up like Mendel's pea plants. It gets read, it gets understood, and it becomes the solid foundation for the next generation of researchers to build upon. So as we wrap up, I want to leave you with a philosophical question to mull over on your own. We spent this entire time talking about the rigid, almost brutal constraints of scientific writing.

23:03If the way we write is so deeply intertwined with how we process scientific truth, does the strict formulaic structure of an academic paper actually limit the kinds of discoveries we are capable of conceptualizing? Or is that exact emotionless recipe the only mechanism we have that makes objective, shared truth possible in an increasingly chaotic world? That is a brilliant question. Without the container, the knowledge simply evaporates. Definitely something to keep in mind the next time you are trying to distill three years of research into a 150 -word abstract.

23:35Thank you so much for joining us on this journey into the art and science of writing up your results. From the Minute Lecture team, we really appreciate you being here. We will catch you next time on the Deep Dive.