In the comments on yesterday’s post, several people asked to see the hips and waist measurements plotted as a scatter graph. The Y-axis (vertical) shows hips, and the x-axis (horizontal) the waist. I really, really, really wanted to see some interesting little clumps develop that might indicate body types- but no. You can see a general trend that as waist sizes increase, hip size also increases. Which is common sense.
This is my favorite graph. I think it’s pretty. The blue line is every individual waist measurement from the smallest I received to the largest. A red square above each point marks the corresponding hip measurement. You can see that on the lower end of the scale, there’s more space between the waist and the hip measurements. This illustrates how the difference between the waist and hip measurements decreases as waist size increases.
This chart demonstrates the relationship between waist and ratio more clearly. The ratio is plotted on the y-axis, and the waist measurements along the x-axis. It shows clearly that the widest range of ratios fall between .7 and .8. I had expected to see somewhat more defined groups. As it is, there’s two exclusive groups: less than .7 ratios at one end of the scale, and greater than .9 at the other.
But really, there’s no groupings of body types easily distinguished by the numbers alone. One can dream, right?
I also looked through the numbers to see if I could find a pattern of relationships between hip circumference and waist to hip ratio. I’m not saying such a pattern doesn’t exist, but it doesn’t seem like it to me. I had expected to find a grouping at the smallest end of the scale around .8-.9 for those who have a petite frame and low body fat, but it seems from the numbers that isn’t the case. Or I don’t have a wide enough sample. I would love to work with more numbers.
Once I played around for a while, I made a sizing chart based on the proportions and numbers from the survey. That’s the first five columns, in both metric and imperial measurement. I left the ratios for the 100cm+ waist circumferences at .86 because while I’m very interested in exploring plus sizing, my data for those sizes is incomplete.
Then I played around and pulled out the ratios for some major (in terms of influence) pattern companies. Those are the first three columns of ratios. Sometimes the sizes didn’t correspond exactly, so I chose to leave those numbers out of the graph. A commentator asked about vintage sizing and ratios, so I dug around for Simplicity sizing charts for the 30′s-60′s. I used Simplicity for the sake of… consistency.
Finally, I asked myself “How much does a tenth of a decimal place in the ratio matter when calculating measurements?” I took the numbers 60-100 (which correspond to the waist measurements) and I took the common ratio for each waist measurement and found the theoretical hip measurement. .71 is the common ratio for a 60-69cm waist, .74 is the common ratio for a 70-79 waist, etc.
Apparently it matters quite a lot. But then I ask myself- does it actually matter when no sizing chart will be able to represent the majority of sizes? I don’t know. Maybe not.
What do you think of those numbers, especially the vintage ones? I haven’t had a lot of time to sit back and think about the implications of these numbers very deeply, so I don’t have much to say about them as yet. Looking through the columns of numbers and deconstructing sizing charts was a great exercise for my brain.
Sometimes I dream about somehow “cracking” sizing and figuring out a system that works well for all kinds of bodies. I think fit is by far the biggest hurdle to new sewists, I wish I could remove it! In my more lucid moments, I can recognize that’s unlikely to happen- there’s too much variety in body shapes.
I have to wonder if it would be possible to create a sizing system using “shape” indicators as well as “size.” I believe Lane Bryant and a few other retailers do this with jeans- how hard would it be to make patterns that way? Something like H (for those with .77-.85+ ratios), S (for those with less than .77 who carry weight toward the back) and X (for those with a ratio of less than .77 who carry weight more toward the sides). The H’s are apparent on the graph, but S and X shapes may have the same measurements but very different shapes…
Or… A sizing system kind of like men’s. They choose pants by their waist and inseam. What if women’s pants could be chosen by waist and hip. “What size are you?” “I’m a 30-40.” “Yes, quite.” I wonder what those patterns would look like?
The more questions I answer, the more questions I find. Don’t be surprised if I bring this up again in a few weeks. But not for a while. Now I want to focus on some hacking!
In the meantime, I would very much appreciate more numbers to work with on my survey. Right now I’m at 502, which is more than I thought I’d get. Could you help me reach a round 1000? Then I can revise.