xaa

Yeah...I saw this about an hour ago, and bit my tongue. But it is dismaying that apparently the reverence for science is so high around here that it extends to downvoting criticisms of science as a process and sociological phenomenon to oblivion.

Maybe it will help these poor fellows if I say look, I have a PhD, I work in research, and I agree with everything krick says. Yes, papers are unnecessarily jargonistic and borderline illegible. Yes, there are massive problems with the incentive structure in science. We openly say science is "publish or perish"; how can we not expect that to incentivize lower paper quality, irreproducibility, and status signalling in the form of unnecessary jargon? Even if we assume the noblest of intentions for every single scientist, which is...idealistic.

spaceseaman is equating (IMO unfairly) criticism of the process of science in its current US manifestation with some kind of disrespect towards its obvious beneficial outcomes and motives. The whole thing IS full of inefficiency, and that's not solely because science is hard. Taxpayers have a right to demand that we don't waste their money and perhaps even to present our findings in a way they can understand with a reasonable application of effort (ideally not paywalled as well).

Ultimately the issue is that we somehow got fixated on the p-value, which (roughly and imprecisely speaking) quantifies the probability that there is any effect, even a small one, rather than using effect size estimates which estimate both the magnitude of the observed effect and the uncertainty in that estimate.

Using p-values as our primary metric means an overemphasis on finding small effects (which are usually not clinically relevant anyway) and unduly low focus on things with big effects.

If an effect is real, but very small, that too may well cause replicability problems because it suggests the effect may not be very robust to small changes in experimental conditions, whereas a big effect would be more likely to be robust.

If you think about the really important scientific findings -- the ones that made a big impact and are indisputably true -- statistics usually aren't necessary to prove them, because the effect size is so large it is simply obvious. I'm not against using statistics anyway, of course, but the point is that we should be looking mainly for effects with big effect sizes if we are after important findings, IMO. It is only a major bonus that big effect sizes are most likely to be replicable.

Because a very common type of hypothesis is along the lines of "gene X is important in disease Y" or "priming people with words will affect psychological outcome Z".

The opposite hypothesis is the null hypothesis which is "gene X is NOT important in disease Y" or "priming DOESN'T affect outcome Z".

Since we assume that most interventions will not affect most outcomes, these are much less surprising and interesting results. They are seen as "water is wet" type of findings and are thus hard to publish because no one is interested.

Now if your hypothesis is something like "X will cause Y to go up" and you actually find it causes Y to go down, that IS publishable. It is only when X has no effect on Y that you will have problems.

Yes, every time I hear people talking about how evil or difficult threads are, I just think back to my frequent use of OpenMP. It is really quite easy to take a single-threaded program and make it multithreaded with OMP -- so long as the jobs only read, and do not write shared state.

Usually I will write a program that does the following:

1. Initialize global read-only data structures

2. Parallelize jobs across STDIN lines (or whatever)

3. Output results within a #pragma omp critical block

It works wonders, and is literally 5 additional lines of code to make a single-threaded program multithreaded with OMP. But only for some types of program. The concept is very similar to what GNU parallel does.

In short, threads are not the problem per se, they are just the wrong tool for the job if you have a multiple-readers multiple-writers situation. Message passing, databases, or something else are more appropriate in those cases. But you will pry my read only shared memory space out of my cold, dead hands.

I have to thank you for your "Ph.D. Grind" memoir. There were many lessons in it, but the lesson I took from it, as I recall from reading it several years ago, was basically that you have to discard perfectionism to get anything done.

I'm not sure if that was even an intended message, but that's what I took from it, and I thought back to it during the dark hours of dissertation writing when I was repeatedly thinking "this isn't good enough". I also remembered the lesson that one should be willing to be flexible in the sense that the dissertation doesn't have to be exactly (or at all) what you originally intended it to be.

Thanks for your interest. I've actually met him once -- a friend of his I was talking to over beers introduced us -- but at the time, he was seemingly more interested in his pending date with the blonde he had just picked up than talking with a lowly graduate student. Can't say I blame him :) His papers are excellent, though. A more philosophical and broad approach is needed in aging, I think. He has mellowed a lot from the exaggerated claims he was making in the early 2000s. Maybe he saw my poster which covered an early version of this work, but we didn't talk about that.

The best aging researcher alive right now IMO though is Jim Kirkland. I've had the good fortune to work a little with him and the man is a living encyclopedia. His brilliance is obvious even in a conference full of PhDs.

Well the project arose like this. I got interested in aging halfway through the PhD. I started collaborating heavily with aging people at my institution. But there are so many papers and so much data to get informed about the area.

So I wondered "is there some semi-empirical way to find out what is 'most important' in aging so I can focus my future efforts on that?"

The solution I hit on was to take all the available gene expression data and to build a system to ask "what genes/pathways/systems change most strongly and consistently with age across species, experimental conditions, and tissues"? This would be a "core aging signature", if it exists. Obviously this is only one of many ways to answer my question and neglects epigenetics, proteomics, etc, although we're currently extending the system to DNA methylation. There is not enough high-throughput proteomics data to make it possible to do this with protein yet. We do not use sequencing for now because it is much more of a processing burden and human RNA is behind dbGaP embargo. And at the time I started this, there really wasn't that much of it compared to GEO.

My boss's interests are much more general than aging, so he encouraged me to develop the system to be more generic while still answering my question, which I did. It became a general meta-analysis system for asking "what genes change expression with <arbitrary condition> across the available experiments in GEO?" We found other things we could do with such a huge amount of expression data, and some of them are in Chapter 5.

I would say the system itself is 80-90% done. But sadly I did not get to a really detailed analysis of aging yet, although my findings so far on that are in Chapter 4.

Not really. My salary was funded by my PI and he was happy to keep someone at my skill level as long as I wanted (and at a bargain salary to boot).

From my department and committee, I started to feel rumblings around mid-year-6 of the "you should get on with it" lines. But they weren't trying to force me out or anything, it was more like concern.

I was spending (and still do) a much bigger percentage of my time on collaborative work than my peers, and they were concerned I wasn't adequately focused on my own career. But that wasn't the case at all -- I was doing what I thought was in my best interests, especially considering my field (bioinformatics) is inherently highly collaborative compared to the wet-lab stuff going on around me.

However, there were some institution-level reasons I cut it off at 7. After that, the institution's policy is that you have to start re-taking some classes you already took (and passed). It was irritating to always have to go to an irrelevant journal club every week (I was in a department that really had nothing to do with my research). Also, I had some appealing opportunities available if I finished when I did. But in no way was I "forced out".

Wow. I just defended a few weeks ago, at 7 years. And honestly, until about year 5, I felt no rush whatsoever to get out. Not that I was derping around -- I finished with 20+ papers -- but I was enjoying it, and I realized that once I was out, two undesirable things would happen:

- I would have to start running the grant treadmill (and indeed, I'm just in the middle of finalizing a training grant that has as one of its requirements...that I must write another grant)

- In the US NIH, there is a 10 year "New Investigator" status that lasts 10 years from the PhD, where you are preferred for certain types of grant awards. So I reasoned it is much better to graduate late and start this clock from a strong position than to rush up the hierarchy and find yourself victim of the Peter Principle.

So for me, delay was a conscious strategy. Only downside was I had to live on a modest stipend. It was worth it for me, though. Another point that I'm making is that "productivity" is not equivalent to "graduating/getting promoted as fast as possible".

As for TFA, I would say that amazing quantities of work can get done by pushing yourself, deadlines, etc, but the creative work that lays groundwork for future growth only occurs during low-stress periods. I try, therefore, to set up alternating periods of both types.

I get why game theory is an attractive model to explain human behavior. But I always wonder why they invent the game, and try to explain behavior in terms of a particular model, rather than using real behavior to fit model parameters, so you could get (pseudo-) empirical numbers for "payoffs" etc.

Or maybe another way of putting it is that clearly rational and human behavior varies with the payoff structure, so it would make sense to include that as another variable.

> I think it is fair to label people inclined to sympathy with Nazi ideas as Nazi sympathizers.

OK, fair enough, but that's not what you originally said, nor is it what people are doing when they throw that term at me. They're (sometimes intentionally) conflating advocacy of free speech for repugnant ideas with advocacy of those ideas themselves. But I appreciate your openness to try to see my position.

Seems like a distinction without a difference. Even if you run a website on your own server, you are still using the ISP's hardware, so they could make the exact same argument. And hosting services are so popular exactly because ISPs have discouraged self-hosting.

Right for service totally without restrictions, no, but "viewpoint-based discrimination" or other types of discrimination are a separate legal category for a reason. "Shirts and shoes required" is different from "whites only" (or, for that matter, "liberals only") on a restaurant door. We allow the former but not the latter, and I think that's reasonable.

> Nothing can, and hopefully never will, stop these groups from running websites

That is literally what OP is about.