Why wouldn't a randomized subsample from 10M people get funded? The required sample size doesn't grow as the population does.

Affects child development. The pattern looks like autism, but nobody's drawn that conclusion; what they have concluded is that electronic devices are more interesting to children than the real world, and cause them to develop emotionally stunted, withdrawn, and more interested in things than people. It's notable you can identify an autistic infant by watching if it's interested in human faces or in objects.

So, yeah. Interactive electronic devices, TVs with robust entertainment content, and so forth draw the attention of children and disrupt their social development. It's believed a similar, but weaker, effect occurs on adults. This is generally framed as "electronic screens are bad", and I don't feel like typing out term papers about what's actually being said because I like to take science for what it is: a pile of important data that must be analyzed for subtle patterns to derive better conclusions on one side, and a simple and complete conclusion useful to engineers but useless to scientists on the other.

You can debate the science if you want, but it's out there, and people have used it to engineer systems of education and general guidelines for the upbringing of children. Such engineered guidelines haven't been scrutinized as scientific principles, but neither has a Boeng 747.

It's more like having the scientific understanding of how lithium ion polymers behave with regard to electron potential creation, and how electrolytic solutions work, and then selecting an electrolytic solution and a lithium ion polymer and putting them together to build a battery. At the end of the day, you've done some work, taken some measurements, made some tweaks, gotten consistent results, patented your Li-Polymer cell, and started manufacturing and selling it in products; it works; but you haven't gotten any science down saying it works the way you believe it works. All you can do is spout about the science that you had for precursor, the things you slapped together, and the results you got.

This might surprise you, but a lot of things are held on the thin branch of slapping a bunch of well-understood science together. Many drug treatments, for example, are held together by science that says certain biochemical effects are useful in a certain way, and science that shows the drug has those effects; we often come back with the conclusion that an entire class of drugs with a long history and variants both ancient and modern are actually totally ineffective because of this.

To put this into context: we have hard science showing that exposing kids to electronic screens is bad. Science backs up that exposing children to electronic screens is bad. We don't have science examining, say, Waldorf Education, which avoids exposing children to electronic screens until they're like 7-8 years old, against new-fangled high-tech Apple Elementary School with iPads all over the place. We've looked at scientific evidence showing that exposure to electronic screens is harmful to child development and determined that a school of education should avoid doing exactly that, in the same way that we've looked at science suggesting antimony should not be in a child's diet in significant quantities and concluded that diets without antimony are better for kids than diets with antimony.

I may be responding to a troll here, but, no, the GP is correct. P-values are about probability. They're often used in the context of evaluating a correlation, but they needn't be. Specifically, p-values specify the probability that the observed statistical result (which may be a correlation) could be a result of random selection of a particularly bad sample. Good sampling techniques can't eliminate the possibility that your random sample just happens to be non-representative, and the p value measures the probability that this has happened. A p value of 0.05 means that there's a 5% chance that your results are bogus in this particular way.

The problem with p values is that they only describe one way that the experiment could have gone wrong, but people interpret them to mean overall confidence -- or, even worse -- significance of the result, when they really only describe confidence that the sample wasn't biased due to bad luck in random sampling. It could have been biased because the sampling methodology wasn't good. I could have been meaningless because it finds an effect which is real, but negligibly small. It be meaningless because the experiment was just badly constructed and didn't measure what it thought it was measuring. There could be lots and lots of other problems.

There's nothing inherently wrong with p values, but people tend to believe they mean far more than they do.

Well my database is on an encrypted server disk in case hackers break in, so I don't need a firewall.

A lot of it is based in cobbled-together science: we know a bunch of things about human development, about psychology, and about impacts of exposure to certain stimuli; we use those to intuit new things. This is basically how new theories are formed, as scientific understanding of two things doesn't necessarily equate to scientific understanding of the effects of plugging those two things together; it does, however, give you a basis for doing so, and a reasonable assumption that outcomes following the predicted model are probably causal.

This is how science starts.

I seriously doubt it. I don't see how location reporting for a payment transaction in which location data is irrelevant could possibly pass Google's privacy policy review process. Collection of data not relevant to the transaction is not generally allowed[*], and if the data in question is personally identifiable (mappable to some specific individual), then a really compelling reason for collection is required, as well as tight internal controls on how the data is managed and who has access. I don't see what could possibly justify it in this case, and I can see a lot of risk in collecting it.

FYI, Google product teams have to develop privacy design docs for all new products, and the designs have to be reviewed by the privacy team (or their delegates) and pass the privacy review before they can be launched. Although Google set these processes up before the FTC settlement, I believe they became part of the consent decree and are now mandated by the FTC and validated in regular audits, so Google can't skip or violate them without potentially-significant consequences.

Disclaimer: I'm not a Google spokesperson and this is not an official statement. It is my personal perspective on the process and requirements. However, I'm a Google engineer who's been involved in launching privacy-sensitive products, so I think my perspective is accurate. I also do security reviews of Google projects, which sometimes touches on privacy issues (though privacy review is separate from security review, as it should be).

[*] Just to head off a likely riposte: No, StreetView Wifi collection and the Safari do-not-track workaround are not counterexamples. They predated the privacy review processes and, as I understand it, were part of the motivation for establishing the processes.

As opposed to what? Cremation? Burial in a box at temperatures well above freezing? You can't seriously argue that this approach makes it less likely that she could be repaired and restarted at some point in the future than typical corpse disposal methods.

It's because of the isolation effect of introducing cell phones and computers to children.

So wait, you fucked up, and now you want us to pay for it?

No, see, here's what happened: School decides they want product X which works with product Y. Product X sucks; product Y is not defective. School has legitimate claim about product X not delivering; product Y is your fault, and you don't go back to the supplier and make them eat the cost.

The school may have a claim against Pearson, since they delivered shoddy, half-ass work. The school has no claims against Apple, since Apple supplied a device not designed to do what the school wanted, and the school intended to extend it with Pearson's product.

There's a real lesson about bad project management and buyer's remorse here; and, looking back, they're ignoring old and proven lessons about not trying to fix education with unrelated technology. The only technology that belongs in education is education: education methods are a technology, and they are the technology for education.

Until you have an education methodology that shows good, scientific basis and utilizes your fancy toys, you're just throwing toys into education. For example: Japan uses a mathematics curriculum teaching students to use complementary number computation techniques, driven by the exemplary platform of a machine called a Soroban; a Soroban would be a ridiculous toy to bring into the classroom if you were not teaching using these computation techniques and trying to leverage the visual and mechanical aspect of learning by soroban (I've done some self-teaching without the soroban, and learned the same techniques; there are, however, scientific reasons to bring a soroban to the table). If they're just doing workbook activities BUT ON AN IPADZ!!!! and not doing anything known to improve education when an iPad is involved, the iPad is a fucking toy not appropriate in education.

It's worth noting there's a school of educational research suggesting that introducing young children to high technology is actively bad, and that high technology should be taught outright after age 10-12 rather than used as a platform to deliver old teaching methods. Small children need most to learn socialization; they need to interact with other children, and not isolate themselves to curriculum. I have my own educational theory which extends this: small children need most to learn techniques of utilizing the brain effectively, set in an environment of free socialization, so as to develop their social behaviors while also giving them tools to rapidly and effectively learn curriculum. In all of these advanced schools of thought, and in mine, you see that pattern: humans need to learn human behavior first, then learn high technology as a tool; wrapping books in fancy electronics won't suddenly make education better.

This is like the 90s when everyone's answer to everything related to computer security was "ENCRYPTION!" Now everyone's answer to every education problem is "COMPUTERS!"