Monday, August 08, 2016


The irreproducibility crisis – an opportunity to make science better

August 8, 2016 by Megan Yu, Plos Blogs

Among the 1,576  surveyed in this news feature, 52% noted that reproducibility is a significant crisis in . Physicists and chemists had the greatest confidence in their respective fields while medical professionals and biologists had the least. In addition, the survey found that 24% and 13% of respondents had published successful and unsuccessful replications, respectively, compared to only 12% and 10% of those whose findings were rejected. These findings are similar to previous studies that found that only 16 of 83 articles recommending the effectiveness of various psychiatric treatments were successfully replicated and that only 36% of replication studies among 100 experimental and correlational studies published in three psychology journals were

Read more at:

Science is what always works.  If it is not reproducible, it is not science.  Any researcher that publishes things that are not reproducible, should be punished according to the list of crimes in the Mikado,( ) or ( )

I often pursue small advances in technology, and sometimes it is hard to tell just what small change in the new prototype is responsible for the change.  Thus, sometimes, I have to go back and retest various aspects of the current technology, to see if I missed something along the way.

Circa 1999, I started researching how to knit warmer objects. It was and is a systematic collection and organization of information about how to knit warmer objects  (aka "science".  I systematically tested the conventional wisdom in the knitting community on how to knit warmly, and found that much of it was/is myth.  I consider this to be no different from published research. I mean, knitters have had centuries to work out how to knit warmly, and to get it wrong is unforgivable.  At the very least, they should be required to sing "Koko" every evening,

Some deny that is is possible to knit "weatherproof" fabrics. That is merely a lack of technique on their part.  I knit weatherproof fabrics.   I have been knitting swatches of weatherproof fabric for about 13 years.  As I learned to use knitting sheaths, I was able to knit entire weatherproof objects.

Part of  the denier's  lack of technique is a failure to use yarns that make the process easy.  Weatherproof fabrics can be knit from a variety of yarns, but knitters are bound up in the mythology of using commercial 1,000 ypp "5-ply gansey" yarns, and these are, in fact, very difficult to knit into weatherproof fabrics.

By, 2006  I was very disappointed in the quality of  the commercial yarns available. and began to spin my own.  However, from the early going, I  found hand spinning also bound up in myth and fairy tails.  Spinners denied the virtues of  differential rotation speed (DRS) and accelerators.  There was nobody around to teach me to spin fine and fast. Sure, there were people like Northernlace, but her approach to spinning fine was arduous. Again the folks using supported spindles  for spinning various fibers into lace, seemed to prefer the slower, supported spindles to the wheel technology.  I understand this because, at those grists, supported spindles are easier than Scotch Tension.  On the other hand, DRS is much more productive than Scotch Tension.

DRS makes spinning fine singles much easier than either of the single drive technologies, and yes, I can use single drive, bobbin lead to spin 56 count wool (~25 micron) at its spin count of 31,000 ypp
( 63 m/g) on AA fliers w/ my bobbins.  So, what? I can also spin that fiber to that grist with a stock Ashford Jumbo Flyer using Scotch Tension! (Using a couple of machine screw washers as tension.) It can be done.   And, I can spin the same singles on the Ashford Lace Flyer.  That is a little faster than the Jumbo.  The point is: I can spin.  My love of DRS is because it is more productive.  Only with DRS can I spin 560 yard/hour of 5,600 ypp and have the hanks come out within 5% of the desired weight (45.4 g)

Today, I have to give more credit to my efforts to damp vibration.   It turns out that with the better vibration damping, I could have run the stock Ashford flyer/bobbin assembly a few hundred rpm faster (e.g., ~2,000 rpm).  However, the extra rpm does not really seem to be useful, so on a practical basis, it was/is the DRS (and AA's little fliers) that boost/ed productivity, and allow/ed use of the higher speed that I get with the accelerator.  And, better vibration damping allowed the higher speeds.  This is particularly true for the Ashford Lace Flyer.  Today, I can run my Ashford Lace Flyer (as a result of better vibration damping)  at 3,000 rpm, but it is still not a practical technology for spinning fine singles in the quantities needed for weaving, or making knitting yarns such as 5-ply sport weight or higher numbers of plies for warmer fabrics.

When you understand spinning, then spinning is faster than knitting.

The first thing that I learned is that "weatherproof" fabric, requires knitting denser fabric.  Knitting denser fabric requires using finer needles and it requires more force to form the stitches.  To retain stretch and elasticity of the fabric, the bars between the stitches are not tighter, but each stitch must be more firmly formed.  Using a knitting pouch is a good first step, but the densest fabrics for the  coldest climates require the use of knitting sheaths and double ended needles.

Warmer yarns tend to be thicker. Warmer yarns tend to have more plies but less ply twist.  Thus, at 1,000 ypp, the 5-ply will be warmer than the 2-ply and the 5-ply with less ply twist will be warmer than the high-ply-twist commercial 'gansey ' that produces a "drafty" fabric.

For very cold climates, yarns in the range of 500 ypp with 10-plies  produce the warmest fabrics with reasonable hand and drape.  Any knit objects intended for the coldest climates will likely have to be knit from such yarns. However, such yarns are not common in the modern commercial market place at this time.  In the current market most, yarns called 10-ply are actually only 750 to 800 ypp, which makes them a 50% less warm than the true 500 ypp 10-ply.

Anyone that says otherwise is trying to publish myth, old wives tales, and commercial nonsese as fact.

If there was a single great failure in this blog, it was thinking in terms of numbers and blurting it out on the blog.  These days most spinners and knitters, do not think in terms of ypp, and thereby much of the content goes right past them.  I use ypp because it makes the math easy.  Wraps per inch (packed to refusal) squared is ypp. The problem here is that the method of taking WPI has changed so that WPI as measured today has no relation to ypp.  I consider the lack of grist, twist, and ply information on yarn bands to make buying yarn like buying a pig in a poke. Likewise the yarn categories are no use to me.  I need to know the number of plies and whether they are woolen or worsted.  This is important if you are designing for a particular climate.  The appropriate yarn differs between SF, Berkeley and  Orinda.  And,  heading to the Matterhorn or Thunderbolt Peak requires an entirely different basis of engineering textiles and objects.  And, that is just in California.  Parts of Vancouver get 12 or 14 feet  (4+ m)  of rain per year.  Then, in places like Fraser, CO it starts to get cold, and we need to apply our our skill to objects intended to be used there.

If we divide ypp by 560 yards in the hank then we have the spin count of the grist.  Then we can buy wool with that spin count, and we can estimate how many yards per pound that wool will spin into. Then we know the total number of yards came be spun from that lot of wool

Anywool can be spun in to 10s (10 worsted hanks per pound, 5,600 ypp, 12.3 g/m) , and it was a standard grist for weaving inexpensive cloth. 10s were ubiquitous and fungible. Everyone knew how to spin them.  At one time 10s were the basis of a good grade of knitting yarn.  It was ubiquitous and fungible. Mostly it was warm enough for a world without central heating, and durable enough to be worn  all the time.  I like such yarns better than what I see at LYS.


Ruth B said...

If your spinning is so super special and superior to commercial yarn, why do you knit almost everything with commercial yarn? Using your own scientific method, we can infer that the performance of your handspun yarn is sub-par or you would have stopped using commercial yarns years ago.

Aaron said...

I knit a lot of my stuff from my handspun.

However, I do not sell my handspun, so I blog about yarns that are commercially available.

I encourage folks to spin, and provide information on how to spin faster and make better yarns.

Aaron said...

I spend a good bit of time knitting, and I think I have bought 10 balls of commercial yarn in the last 5 years.

Moreover, one MUST compare the performance of one's hand spun to the performance of the best commercial yarns. Thus,one must continue knitting commercial yarns.

Circa 2010, I had some bins of commercial yarn with more than 10 pounds in them. Most of this is gone. However, I keep ~3 pounds of commercial 840 ypp, 6-strand cable in case patknitter ever wants to knit a serious yarn that is willing to stand up and fight back.

knitphomaniac said...

"However, I do not sell my handspun, so I blog about yarns that are commercially available."

... so? If you're going to bother blogging (and devoting a whole blog) about spinning, and knitwear with spinning, the knitwear you show off should revolve around your knitwear from your spinning. I've seen more commercial yarns knit in your blog than I've seen spun yarn, so it's a bit hard for me to believe that you've only bought 10 balls of commercial yarns, and yet you supposedly have bolts and bolts of spun yarn.

Let's call a "spade" a spade shall we, hm? ;) Your blog isn't about spinning or knitting, it's about how to build tools to spin. Your "expertise" is in making fancy tools, not better yarn.

Aaron said...

The difference between us is that when I am shown a different way, I test it, and if I like it, I grab it! If I do not like the new tool/skill/material, I go back to the best way that I know. Every craft is a tripod of: tools, skills, and materials. If tools were not important, there would not be so many knitting needles offered for sale!

When you are shown something different, You say, "Can't be!" or something else depreciating.

The result is that I learn new stuff, and you do not. It does not matter if the new stuff is a tool or skill or material, if it makes my knitting better, I grab it and test it, sometimes for years. Your failure to learn as the result of a closed mind is not my problem.

knitphomaniac said...

" If tools were not important, there would not be so many knitting needles offered for sale!" - um, what?

Knitting needles are offered for sale because the market demands it, and inferior tools is not the reason why - but it's amusing that this is the only factor you seem to focus on. I have knitting needles that I've inherited from my grandmother, I have knitting needles that I've also bought (once, per size) decades ago. Some are wood needles, some are metal, some are plastic - and which needle I use depends largely on the material I'm knitting with.

The only reason why most people re-buy the same needles is largely due to either wanting to make more than one project using the same size needle OR losing a pair, it has nothing to do with the quality of needle itself.

Yet another thing you're wrong about. :/

Aaron said...

I would say that different kinds of yarns (woolen, worsted, multi-ply worsted with low ply twist,etc.) demand different kinds of needles.
Swaving demands its own curved needles. Dutch knitting sticks demand shorter, stiffer needles which also work well with goosewing knitting sheaths. Knitting (adult) sweaters from commercial 5-ply sport weight demands long flexible needles (with a knitting sheath.) Blunter sweater needles will allow knitting 10% faster, which can save almost a day per sweater. However some pattern stitches need more pointy needles, which slow you down on the plain knitting. Of course, I can knit an adult sweater on shorter needles, but it is ~20% slower - that is 2 days (16 hours)longer than my baseline in the knitting of an ordinary sweater. Thus, the difference between knitting a sweater with long blunt needles and knitting with short needles is 3 days (24 hours), that is almost a 25% difference in knitting time depending on the kind of needles that I use. On the other hand, the short needles are more convenient in some places.

I have as many needles as I need.