schuetze

If you are reading popular books on social science (which many business books are), it is very likely that the author is pushing research they are personally invested in succeeding. They are often not trying to convey a nuanced understanding of a phenomenon. Furthermore, popular books are often not peer-reviewed to the same extant or at all as regular research literature.

A better alternative is to read review articles in journals. They are shorter than books and will also give a more reasonable estimate of the certainty experts have in a certain field of results. For example, in psychology there is the Annual Review of Psychology, which generally publishes ~50pg readable summaries of literature by experts in the field. For the love of social science, please stop reading popular books on the brain/mind.

As a psychologist/cognitive scientist (my identity depends on my mood/day), I was equally disappointed with the sections on psychology and neuroscience. Some of the questions have been answered (at least partially, such as the importance of domain knowledge versus "processing power").

Overall, it was a very odd look into a discipline that I am familiar with, which leads me to think that the ideas are not very promising for any of the fields I don't have expertise in.

I went back through the press release sources and found that it's actually based off of two recent papers:

[1] https://academic.oup.com/cercor/advance-article-abstract/doi...

[2] https://www.jneurosci.org/content/early/2020/03/30/JNEUROSCI...

Perhaps this is aimed at non-American healthcare systems, but I have not seen a single report of American hospitals running out of ventilators, let alone BiPaP machines. In order for this to be a responsible hack — used under the guidance of medical professionals — we would need to have run out of both.

If anything, at least in the American context, I'm more worried about running out of the sedative necessary for ventilation.[1]

I have nothing against hardware hacking in non-pandemics. If you want to hack your own insulin pump or create epi-pens on your own (non-crisis) time, that's fine by me.

But I think the cost-benefit-risk analysis changes in pandemics, because people are too hungry for easy fixes and make ill-advised decisions under pressure. For example, even doctors (ostensibly medically-literate professionals) are prescribing themselves hydrochloroquine [2], which does not seem to be a miracle cure and sometimes, itself, dangerous (and also leaves lupus sufferers at risk of a disrupted supply chain).

[1] https://www.vox.com/2020/4/6/21209589/coronavirus-medicine-v...

[2] https://www.nytimes.com/2020/04/12/health/chloroquine-corona...

Agreed. BiPaP to ventilator already seemed like a somewhat questionable transformation, and now we're going

CPAP ----questionable software hackery ----> BiPaP --- questionable hardware hacking ---> Ventilator

This is hubris. Ventilators are not iPhones circa 2010. It's irresponsible for non-medical researchers to not only pursue, but also disseminate, these jailbreaks. A significant portion of medical device RnD is related to creating technology that is hard to misuse and won't result in accidental death, and I just don't see that here.

From personal experience as a cognitive psychologist, I have found it prohibitively complicated (albeit possible) to implement randomly counterbalanced experiments in Qualtrics. Stimuli sampling and reaction time measurement are particularly hard to implement (unless they've added more functionality recently).

All three of these pieces of functionality are required for many common paradigms in experimental psychology.

Although the time to completion seems quite disparate between Europe and US, the difference isn't nearly as stark as it appears.

Nearly all Europeans enter PhD programs with a Master's degree in hand. At least in psychology (my discipline), US PhD programs will admit candidates directly from undergraduate. The overall time from bachelor's to PhD is similar when you account for these differences.

It also seems that Dropbox Plus somewhat recently gained "smart sync" capabilities, allowing for the storage of selected folders/files in the cloud only.

As a meta-recommendation, I suggest checking out Five Books [1]. It's a website dedicated to bringing in experts and having them suggest five books that best represent their given fields. The archive of interviews on Five Books covers all the topics listed above and more.

[1] https://fivebooks.com/

Thinking in Systems: A Primer by Donella Meadows

Once you start looking for positive and negative feedback loops in the world around you, it's hard to stop. In particular, Meadow's book is great because it also goes beyond +/- loops in isolation, and shows more complicated patterns, such as eroding goal patterns and traps that often cause public policy interventions to fail.

Do you have any evidence for these statements?

In my experience, it's generally the budget-strapped universities that choose to replace tenure-track professors with adjuncts.

On the other hand, private universities with very large endowments tend to be prestigious and part of maintaining this prestige means hiring tenure-track professors to teach undergraduates, mentor graduate students, and to do research.

Have you tried accessing the publications via the on-campus computers in the normal undergraduate libraries? I don't think you need to be in a medical building to have access to the online medical journals.

Well there was, but it was taken down by Daniel Webster in front of the US Supreme Court.

It’s a surprisingly common mistake though. I think I see Dartmouth University more often than Dartmouth College when reading non-Dartmouth materials.

I use TiddlyWiki for note taking, and I think it has most everything you’re asking for. Except, my one complaint is that it uses an odd dialect of markdown.

The replicability of the biomedical sciences is just as bad if not worse than in psychology.

[1] http://www.slate.com/articles/health_and_science/future_tens...

[2] https://www.nature.com/news/1-500-scientists-lift-the-lid-on...

You can also look into "system dynamics" software, which allows for the modeling of complex systems with very little user-facing mathematics. Dana Meadows wrote a book called "Thinking in Systems," which is a great intro to this subject.¹

I personally have used Stella by isee systems, which will output an equation from your model if you are so inclined, but I think there is cheaper/free software out there that will do similar types of modeling.

1. https://wtf.tw/ref/meadows.pdf

For my undergraduate thesis, I researched the science of learning. Many of the ideas put forth by this article have merit, but I would also recommend searching the keywords of "desirable difficulties" and "retrieval practice" (self testing) in Google Scholar as potential avenues for examples of more concrete study strategies.

Research shows that the best way to learn is to challenge oneself. Self-testing, in particular, is a highly effective way of avoiding the illusion of fluency (Long-term Learning != Current Performance) and creating durable memories.

Further reading:

1. Make it Stick

2. "Making Things Hard on Yourself, But in a Good Way: Creating Desirable Difficulties to Enhance Learning" by Bjork & Bjork [https://teaching.yale-nus.edu.sg/wp-content/uploads/sites/25...]

3. For a survey of student misconceptions about learning: "Instructor and student knowledge of study strategies" by Morehead et al.

It works as long as no one takes a screenshot.

Under non-competitive circumstances this is true. However people are actually very adept at generating nearly perfectly random choices in competitive games. There is evolutionary pressure to produce randomness, because if you are predictable in a game, you will lose.

Citation: https://www.sciencedirect.com/science/article/pii/S089360800...

Psych Science, while published by a somewhat reputable source (the APA has had some questionable positions on torture), is not considered a highly rigorous journal. It prioritizes “sexy” findings that are more likely to not replicate. Within the psych community, these kind of findings are taken with a grain of salt.

“Thinking in Systems: A Primer” by Dana Meadows. One of the best books for understanding the worlds systems intuitively. Allows you to use a mathematical framework without needing precise numbers to get a model off the ground.

Many social scientists are adopting Bayesian methods. However, they don't always have much of a background in college math. These kinds of guides afford them the intuition necessary to begin implementing related methods in their research.

In some ways, I think the unintended consequences of these ads are very positive. Anything that encourages us to think about the ever-increasing amount of data held by private companies is a good thing in my mind.

I doubt that we will ever break the grip of the advertisement-driven web, but maybe someday a corporation will make an advertisement tone-deaf enough to spark a revolution on how the internet is funded. I'm not affiliated with, nor do I have any stake in this concept, but I think micro-payment funded content, like yours.org, is an interesting alternative. But I doubt many people will adopt yours.org, myself included, until the content is on part with that of ad-funded media.

Not OP, but I'm personally a fan of chill-hop stations, such as Lo-fi Hiphop Radio 24/7. This kind of music lends itself to studying and working, as it has very little lyrical content. Find it here: https://www.youtube.com/watch?v=AQBh9soLSkI

Step 1: Become Bayesian.

Seriously though, I am inclined towards approaches, such as pre-registration, which limit the number of researcher degrees of freedom in analysis. It's not necessarily that statistics are broken. It's that the system incentivizes researchers to break the assumptions underlying these statistical tests.

IMHO anonymity has never been Bitcoin's strength, and we must stop pretending that BTC is the right tool for transactions that need to be private. At best, Bitcoin is pseudonymous.

Other currencies have attempted to fix this problem, such as Zcash. But I think it will be hard to escape the volumes of metadata created through transactions on merchant websites. Ultimately, your spending habits and browser cookies will say more about you than your BTC address.

It looks like DSL's website hasn't been updated since I found out about it 10 years ago.

Considering the relative ease of implementation, classification accuracy with smaller datasets, and computational efficiency of Naive Bayes classifiers, I am surprised that they are not mentioned as often as other machine learning competitors, such as random forest.

Are there major drawbacks to Naive Bayes classifiers? Is it just that they aren't as accurate on large datasets?