Jordi Prats studied Biology at the University of Barcelona, where his interest for statistics was born. He then obtained my PhD at the Civil Engineering School of the Polytechnic University of Catalonia before becoming an environmental modeler, specializing in Hydrology, Freshwater Ecology, and Lake Hydrodynamics. He has worked in the public and the private sectors in France and Spain and has collaborated with Significance since 2011 where he is now a member of the editorial board.
Episode Description
For some people, shopping for clothes can feel like a step into Dante’s Inferno, all suffering and punishment. Even for those that do enjoy shopping it can hard to find something that fits well. The data for size is the focus of this episode of Stats+Stories with guest Jordi Prats who recently wrote an article about the issue in Significance Magazine.
Check out his article in Significance now!
+Full Transcript
Rosemary Pennington
For some people, shopping for clothes can feel like a step into Dante's Inferno, all suffering and punishment. Even for those who enjoy the activity, it can be difficult to find clothes that fit well, with sizing seeming to be inconsistent. The data for size is the focus of this episode of Stats and Stories, where we explore the statistics behind the stories and the stories behind the statistics. I'm Rosemary Pennington. Stats and Stories is a production of Miami University's Department of Statistics and Media, Journalism and Film, as well as the American Statistical Association. Joining me is regular panelist John Bailer, emeritus professor of statistics at Miami University. Our guest today is Jordi Prats. Prats studied Biology at the University of Barcelona, where his interest for statistics was born. He then obtained his PhD at the Civil Engineering School of the Polytechnic University of Catalonia before becoming an environmental modeler specializing in hydrology, freshwater ecology, and Lake hydrodynamics. He's worked in the public and private sectors in France and Spain, and has collaborated with Significance since 2011, where he is now a member of the editorial board. He's written a two part series for Significance about why it's so hard to find clothes that fit. Jordi, thank you so much for joining us today.
Jordi Prats Well, thank you for having me. It's, it's a great honor to be here
Rosemary Pennington How did you get interested in this issue of the fit of clothes?
Jordi Prats
Well, it was something that actually happened. It all started because my partner told me about this nice TV show, onBBC, the Great British Sewing Bee. Now if you know it, yeah. And well, they started looking, watching the show, seeing how they made it, what they wanted to do, too. And I started to investigate the demand. And at a certain point, I started taking classes on pattern making and sewing. So it was a bit of an extension of my area of expertise.
John Bailer
That's a great start. I wasn't guessing this was the path, right? Yeah, this is really, really interesting. By the way, I'm still thinking about Rosemary's comment about trying on clothes being like going to Dante’s Inferno, a vision of hell, it's terrible. This is, you know, as you went through this article, one thing that really jumped out to me was your reporting of a 2008 study that said that 40% of Spanish women had problems finding fitting garments. And they were, you know, mainly because they're too small was what you had reported. And that you're saying that other studies have suggested that clothing, that ready to wear clothing, is designed to fit only 65 to 85% kit, can you talk a little bit about some of the you know, how did you start digging into the these data and you know, come up with some of these these reported values?
Jordi Prats
Well, it was a long search actually. First, I started looking for pattern to pattern making manuals, I looked in the past and incentives and things back to the 19th century. They see what it is that people were doing and how these evolved. And at a certain point, I found that there were new technologies to take measurements of people: 3D measurements, to 3D scanning tools, where people wear a garment, covering the body and then the measurements of different parts are reconstructed to see the 3D shape. And there have been many studies of this type in the UK, in the USA, in Spain, too. And to study, the 2008 study is one of these, they measured 1000s of women, and they also made the authority the task of a series of questions in relation with the measurements. And one of the results was that many, many women had problems finding a fitting garment. But one should say that it's not exclusive to women, this is prominent, and it happens to men too.
Rosemary Pennington
So how do manufacturers figure out what the sort of standards around their sizing is going to be? How do they decide, like, this is a medium, this is a large? What is the process? And what did you learn about that process as you were working on this article?
Jordi Prats
What you learn is that they usually have a public, a target audience. I looked at different brands, and also a different pattern maker, magazines, you know, magazines that sell patterns to make clothes. And it noticed that they have different objects, some, for example, in magazines, they found that there were some of them who catered more for the average woman, while there were other ones that were more directed to slim women. And however, what I didn't find is magazines or brands that were dedicated to people with unusual measurements, larger sizes. In fact, I started looking at this some years ago, and one of the brands I looked at didn't have large sizes last year. But this year, before publishing the article, they had this. There has been an evolution to it, to include larger sizes. But they know they don't arrive to the most unusual, either, because they are too large, too large on one side or too small on the other.
John Bailer
So I'd like to have you just tell us a little bit about the history here. I mean, that's one thing I really liked about this, this Significance paper of yours. And this idea, that there was this ready to wear idea of clothing manufacturers that emerged with the idea of targeting certain sizes, and then what's called pattern grading. So can you just just give us a little bit of a history lesson?
Jordi Prats
Yeah. Well, the beginning tailors, well, just made close to the measurements of the clients and then the demand of the growth to feed them. However, sometimes it happened that they had requested to come by mail, or the client was not there. So they started devising systems so that with only a measurement, the girlfriends of the chest, they could still throw the few different patterns for the four different guns. Now the problem is that there's a lot of variability in the human body. And what they did this while there is what is good at the moment is just use a proportionality rule. Assume that if SQL fences increase in a certain proportion, all the other measurements would increase in the same proportion. Now, this is to pool for folly to proportion of the population but not for everyone. And like this, they knew that it didn't work every time; they knew that there were problems and during all the 19th century manual after manual of pattern making that says no we do it better; we do it more scientifically, methods are more geometrical, etc. But they did just little, little improvements, say we take an extra measurement here, or we use a different constant of proportionality. So in the end, in the beginning of the 20th century, it was still a problem. And it was when there were the first systematic measurements to obtain data in order to be able to develop a more rational sizing system. And it was, in fact, at the First World War, when soldiers were coming back to the United States, many of them were measured. So that the army could know which, which was the liability in the in demand preparation. And so they could prepare the sizes for soldiers. Now this was made for men, but for women, they did not study until many years later. And the same for children. But then, there was a study by O'Brien. In fact, there were two studies, one for children and one for women. They also took a lot, a large number of women and children, to take measurements, and then they analyzed statistically the data to see which were the measurements that best describes how we get the body size, in case.
John Bailer
So we saw this, there was an evolution from tailor made kind of made the order to proportionality rules that weren't quite there. And then finally, someone started to measure some data on body parts and relationships to body parts, that kind of informed this, and the fact that this was kind of inspired by needing to mass produce uniforms for armies, that seemed to drive a lot of this early on.
Jordi Prats
Yeah, in fact, the industry still didn't have these requests for ready made garments, they didn't start doing this at scale. So, for example, if it started a bit earlier with after the Industrial Revolution, when they were the first department stores, to which in fact, it was some of the first examples of fast fashion also, fast and cheap fashion, because they needed to sell lots of product at a fast pace. So this is something that a bit popularized the idea of buying ready made instead of making too much measure, and, and yeah, and, and it is important also for the army, because if you have to work in straining conditions, you need to do it well, to take off comfortably.
Rosemary Pennington
You're listening to Stats and Stories. And today we're talking with Jordi Prats about statistics. The point you make about sort of this being an early form of fast fashion, made me sort of zoom to the present moment and think about all of these different brands and department stores that have sort of popped up, that are designed to cater to sort of fast fashion tastes, you know, online, there's like Shein, which is a retailer, there's H&M, and other stores like that Primark, which are producing these sort of articles of clothes that are meant to sort of appeal to a mass audience, but also, in the hopes that, you know, in a year from now, you'll be back buying something else, right, like that turnover. The interesting thing that I have found in visiting these places is that the sizing between these places is never the same. So I might be one size in one place and then one size in another. And I wonder, as you were doing your research into this issue of sizing, what did you find around this inconsistency in sizing between stores? Why is there not a standard? Like this is what one this particular size means?
Jordi Prats
Well, in fact, there are some standards. The problem is the industry doesn't use it.
Rosemary Pennington
Oh, well.
Jordi Prats
You affect the blind study, it was studied, people pose a standard measurement, standard for the US, and it was proposed, and several standards were proposed afterwards. But since it's not an obligation that each maker uses the sizing grades, the more appropriate for the target preparation. And then this concept is known as vanity sizing. So that one make can get more, can sell more clothes, because the selling here, a smaller size, but which in reality is not small, but it's the same. So this sort of competitive adaptation makes them tweak the sizes too, to make them appear more, more desirable.
Rosemary Pennington
That's interesting because for years, I was the same size, even though my body changed over time, and I was like, This can't be right, but I'm gonna keep buying these clothes.
John Bailer
As you're talking about this, I try to think about whether my own vanity sizing has applied. I see if that is something I've experienced, looking for clothes. You know, so one of the things as I was looking through your article, you're talking about how the proportionality method was in part rejected because of regression, with addressing this and that, you know, there's this movement of moving from kind of a single variable that might characterize a sizing strategy for ready to made close to milk to a couple of variables based on, you know, what you're providing kind of uncorrelated but important information. Can you talk a little bit about, you know, how that evolved?
Jordi Prats
Yeah, so, it was adopted by the cutterhead, who made the first systematic studies of the measurements of the human body. And, thanks to these studies, they proposed the par impact studies, they proposed the normal distribution, the fact that the measurements of a body, they follow up, sort of Bella distribution, where they said that the ideal size would fit the one that was in the middle of the average, while those that were farther from the center were due to errors in God's plan, or what they would say. So, there was the first distinction that was the average, the most important. However, they started also looking at the relation between the body parts. And they saw that, well, they didn't increase at the same rate as they would increase according to a proportionality right. Later. Galton, he also saw, he said more, in depth distinct. In fact, he was studying a couple of questions related to forensics, and also to turn to topology because when anthropologists discovered a body, they didn't have all the different bodies they had, maybe a bone from desire or whatever. And they hate to estimate helpful, it was called ODHH. And then he wondered if knowing just one measurement, one could estimate the rest of the measurements. That was the bit of the deity that made the thing off correlation.
John Bailer
So I'm curious as you worked on this, what was the biggest surprise for you?
Jordi Prats
Well, maybe the fact that year after year, the in the 19th century, dealers continued saying our method is better, our method is more mathematical, more rational, and in the end, it didn't improve very much, and also the fact that up to this moment, the situation hasn't improved that that much even part okay. Because there is a lot of tending to always use the same method and not tending to change them. But there has been another phenomenon that has made the province wars which is at the same time that the pattern making methods slowly evolved. The size of the measurements of population also increases with time. So we tend to be taller or we tend to be bigger, we tend to be more diverse, too, then there is a mix of several populations. So we have much more viability now than there was when tailors used the proportionality methods 200 years ago.
Rosemary Pennington
I love, in the article, where you mentioned Gulliver's Travels. That was one of my favorite stories as a kid. So I just really was like, Oh, yes, I remember that scene from the book. I wonder, you mentioned a little earlier that there are the standards that exist that retailers don't want to use, because they can sell more if they can sort of outside them. I wonder, given sort of this discussion that has sort of bubbled up in the fashion community around fast fashion, the problems associated with it, as well as this problem of sizing and fit, whether there is, in the work that you've done, whether you found any evidence that maybe there's a desire to move towards some more standard size in order to not turn away customers, right, because it does feel like frustration around clothing, shopping for clothes, often is centered on this, this inability to find clothes that fit.
Jordi Prats
What they have seen is that that is a request for a wider range of sizes, that's obvious. I have also seen that there are several brands that tend to increase the range of sizes. There are also some bands that are dedicated only to large sizes. But I can also see that, let's say, the average behavior is like being saved. So you, you make your garments for the center of the distribution, why you know, that it's, it's more difficult to go wrong. Because there is less material. When you go far, far from it, you don't know why, it may be that your shoulders are wider than the average. Or it may be that your hips are wider than the average. Or it may, it may be that your request's major measurement is wider than the average. Or it may be because you're tall or short. There's a lot of reasons that can make a garment not just fit. So it's much easier to make a garment fit when you are near the center of the distribution than that, then when you're apart.
John Bailer
In your paper, as you're finishing it up, you talk about the problem of dimensions. And an aspect of this, you said that the problem that really, fitting comes down to identifying representing the human body in some model in terms of some variables or dimensions, you know, and this minimum number of dimensions, that kind of captures as much of the variability in the population you're trying to model, granted that the human population is probably a mixture of different populations, but for the sake of this question, if you're focusing on a single population, what do you think the future is of trying to capture this model for human size in the smallest number of variables that can be used to represent and then provide a model for fitting?
Jordi Prats
Well, in fact, part of this is a work on my second paper on the subject which has not appeared yet.
Rosemary Pennington
Yeah, John is trying to give too much away.
John Bailer
So, we got to have you, Mr. Crowder, for a return visit.
Jordi Prats
Well, I'd be happy to. In fact, one possibility is quite obvious: we will be using principal component analysis. So you can select which are the deeper levels that explain the most of the viability, and well, I have seen a master study that did something like this, and I am sure that whichever brand that is interested, has already done it. For example, some years ago, you thought it was a lovely thing to make measurements of many women also using 3D scanning, and the proposed and more appropriate sizing systems. But then they do also identify nine types of body, chest, chest morphology. And for each of them, they could recommend different types of longevity. So that there are some things that are being done. And then there's another thing, although it's not, I think it's not very common, which is already selling things adapted to your own sizes. So if you can measure, you know, it can give you your measurements in some way. Say for example, there is an app where you can take a photo, or two photos, and then it has made you your measurements. So you can send these to the production center, and then make the garments just adapted to yourself. It takes more time, taking more time to get the blood, but that you use less time and go search off of your garment. And that would also reduce unsold gallons, because you only produce what you would use. So that's a possible future.
Rosemary Pennington
Well, that's all the time we have for this episode of Stats and Stories. Jordi, thank you so much for joining us today.
Jordi Prats
Thank you. Happy to be here. Great fun.
Rosemary Pennington
Stats and Stories is a partnership between Miami University’s Departments of Statistics, and Media, Journalism and Film, and the American Statistical Association. You can follow us on Twitter, Apple podcasts, or other places you can find podcasts. If you’d like to share your thoughts on the program send your email to statsandstories@miamioh.edu or check us out at statsandstories.net, and be sure to listen for future editions of Stats and Stories, where we discuss the statistics behind the stories and the stories behind the statistics.