rflrob

7 ms·

1.
by rflrob 2mo ago
There are always going to be people whose reason is “because we can”. Going to Mars has, on the scale of a human lifetime, only recently become vaguely feasible. As someone who works in STEM, I’m generally on the side of new technologies un
2.
by rflrob 3mo ago
More data sounds better, but especially in a medical context, you have to be careful, because false positives have consequences. The PSA test is no longer broadly recommended for prostate cancer screening [1]. What harm could it do, you k
3.
by rflrob 6mo ago
Stearates aren’t microplastic plastics, though, they’re just similar enough under a microscope and in some chemical analyses. Without knowing which stearates glove manufacturers use (or what exactly it is about microplastics that is harmful
4.
by rflrob 10mo ago
I’m curious how you came to that conclusion. While he’s certainly not in the pantheon of best presidents, he ends up around the 75th percentile in rankings by historians. Even subtracting a few spots, he’s nowhere near close to “one of the
5.
by rflrob 10mo ago
I don’t see a straightforward way this would actually help with the cleanup. A hypothetical microbe that “eats” oil would be useful in an oil spill as would chemically break down the oil and harvest its carbon. A radiotropic fungus that’s i
6.
by rflrob 1y ago
To tag on, the economy gives us more than just physical "stuff" (though that's one of the easiest things to measure). Do you enjoy travel? How about clean water, yoga classes, or therapy? What about the ability to post a mes
7.
by rflrob 1y ago
There’s very often a comparison to the somatic (i.e. non-cancer) genome of the same patient. It’s a great way to quality control that there wasn’t some sample mixup in the lab. Transmission of cancer is rare in humans—if it were not, it wou
8.
by rflrob 2y ago
That assumes that what causes the false positive is some kind of analytical noise in the test. The bigger concern is biological noise that would persist if you tested the patient again. It might still be useful to know you have weird protea
9.
by rflrob 2y ago
> or worse placebo Just to be clear, most drug trials for anything where we have an effective treatment are not “new drug vs placebo”, but instead “new drug vs standard of care”. Thus the goal being to prove it’s better than what already
10.
by rflrob 2y ago
I always hesitate to ascribe motives to non-human animals, but the butterfly shouldn’t particularly “care” whether the pollen gets from its body back to a plant. If the butterfly is eating the pollen, then maybe there’s an advantage to hoov
11.
by rflrob 2y ago
I’m not OP, but the startup ethos of the last few decades has been to move fast and break things, fake it til you make it, and generally operate on the bleeding edge to deliver incredible products (and, occasionally non-credible, cough Th
12.
by rflrob 2y ago
Also the Guatemalan coup that was heavily lobbied for by the United Fruit Company. https://en.m.wikipedia.org/wiki/1954_Guatemalan_coup_d%2527%...
13.
by rflrob 2y ago
The claim, as I understand it, is that commerce within a megaregion is more tightly integrated than commerce between megaregions. If this is the case, there should be more trade between Baltimore and Boston than Baltimore and Raleigh (contr
14.
by rflrob 3y ago
While I agree that you want to be wary of any particular claims, and more study is warranted to establish the effectiveness of any particular intervention based on data, there’s a Bayesian justification to looking at otherwise unexplained d
15.
by rflrob 3y ago
In addition to the comments proposing some other kind of exercise as a control group, I’d want to see whether “being prescribed something but then being waitlisted“ prolongs depressive symptoms compared to, say, “being prescribed and receiv
16.
by rflrob 3y ago
In addition to the sister comment (that I 100% agree with) that publicly funded research ought to be in the public domain, if the technique is broadly useful I’m sure that companies can find patentable optimizations to the equipment or prot
17.
by rflrob 3y ago
23andMe works by checking ~1M sites along the genome, and using known correlations with pathogenic variants. This Invitae test actually sequences the (coding portion plus a bit of) the genes involved, and can detect rare[0] mutations. Addit
18.
by rflrob 3y ago
While the Linear, No Threshold model is relatively widely used by regulatory and safety bodies, evidence for it (or any other competing mode of low dose exposure) is quite weak. https://en.m.wikipedia.org/wiki/Linear_no
19.
by rflrob 3y ago
chr6:41,515,652G>C, GRCh38, rs9367106 Is the lead variant in the association study, but that doesn’t mean that it’s the causal variant. Genome Wide Association Studies line the one described here are relatively blunt instruments that are
20.
by rflrob 3y ago
The article is from a .au TLD, and the conversion to Australian dollars is roughly correct.
21.
by rflrob 3y ago
You may be pleased to know that “essay” literally means “attempt”. The idea being that you aren’t sure exactly what you believe, so you’re going to attempt to write it down and see if that helps to clarify it for yourself and others.
22.
by rflrob 3y ago
The introduction (but nowhere in the rest of the article) makes a distinction between Dollar Cost Averaging (where you spread out a windfall over some period of time) and systematic investing (SI; where you regularly invest income as you ge
23.
by rflrob 3y ago
One thing that’s confusing is that Nature has two purposes: first as a scientific journal, and second as a science news magazine. They’re bundled in the same physical issue (though there are also branches of the journal, eg Nature Genetics,
24.
by rflrob 3y ago
I think the biomedical academic job market is tight enough that you could fairly easily attract a strong class of researchers to nearly any second tier city. Maybe not the best of the best of the best (at first), but there’s enough bright p
25.
by rflrob 3y ago
Bonobos are exactly as close to us as chimpanzees (by descent). Maybe the “default“ human behavior, independent of the last x thousand years of culture, is closer to one or the other, but I don’t think there’s a good a priori reason to assu
26.
by rflrob 3y ago
Depending on the audience (and here I’m assuming mostly tech folks in t-shirts), it’s just as possible that you might not be wearing anything woven, and instead entirely clad in knitted fabrics. I’m by no means an archaeological technologis
27.
by rflrob 3y ago
Whoops! I originally had “does X”, then changed it to make it less abstract.
28.
by rflrob 3y ago
For some scientific disciplines, code (software) is not a mass noun. “I wrote a code that does processes the data files” is acceptable, even though it sounds wrong to my ears.
29.
by rflrob 4y ago
One obvious difference between how the AVSAB and SAT are used is that the minimum score on the AVSAB is ~ 35th percentile (somewhere around a 980 on the SAT). That kind of score is likely all-but disqualifying at Columbia, where the 25th pe
30.
by rflrob 4y ago
> if you skip classes in college, they just shrug, cash your tuition check, and give you an F. Depends on the college. A friend of mine who went to an “elite institution“ struggled to supervise themselves and got set up with a counselor

More ›