I am sure I will get some kickback on this one, but Tynker has been great for my 8 year old. It is basically a Scratch 1.4 clone with storyline, challenges and specific lessons/projects.

I'm kinda torn about it as it takes a free product and commercializes it, giving only some lip service to it. However, I figure the storyline and projects and badges system has some added value, and I haven't found similar interest building services available for Scratch. Though if you know of any, please let me know.

As an added advantage, he'll easily be able to transition to Scratch and take advantage of the remixing which I consider to be one of the strengths of Scratch.

I'll preface this by saying I don't disagree, medhelp.org does do a poor job if their intent was to display a scientifically rigorous and reproducible statistical analysis. The OP may have been able to find a better citation to support their argument.

But... medhelp.org does say

I would offer that would at least they are pulling these numbers from the total mortality rates. So you have a 1 in 800 lifetime chance of dying from a motorcycle accident. Probably somewhat higher if you were between the ages of 16-35, or even higher if you actually rode a motorcycle. However, those are also educated guesses.

I didn't say that the article claimed a 1:800 per year chance here, instead I pulled this stat from the article:

Which by your best guess calculation was actually not far off, if you were calculating it for a full year.

Those numbers probably have nothing to do with the 1:800 calculation and is instead presumably a lifetime risk estimate. So the statement is entirely absurd, but primarily because it is likely accurate.

Your calculation has a few errors. You are estimating on a per ride basis and making a number of assumptions in your calculation, without references to support those assumptions. The stats are an annual thing, according to the citation, which 650 000 / 52 weeks/year = a little less than 13 000 - pretty close - within an order of magnitude - to the 4762 cited.

However, this is not the best tool for the job at any rate.

Check out micromort - which estimates for every 6 miles travelled by motorcycle you have a 1:1 000 000 chance of death. This is a much better unit of comparison. Sadly the reference from there is paywalled.

Yes, motorcycles are bad. Add in disability and you really would be stunned.

Like how exactly does a mountain made of iron burn?

Great, so what to do for the 4-6 band kid who does well with reading and comprehension, easily meeting or surpassing the next band, not so hot in math and not meeting the requirements for the next band, but is young or socially young, but can Riverdance with the best of them and still draws stick figures?. Which band does s/he go to? Why are 4 bands better than 12 grades?

I will resist making fun of the provided magic thinking here that seems to be right out of someone's feelings on what might work well. We could spend quite a few years going down every garden path on a whim to see it it might just be better.

1. What is the goal?
2. What is the metric to measure that goal? Is it valid?
3. What paths have been tried and can we now retrospectively measure those ones with the given metric?
4. Take the retrospective methods that score well and do a prospective cohort study to determine where to go to.

Quite frankly, until we can even agree on step 1, we will get nowhere with an overhaul of the school system.

We've got plenty here, lets send it along.

I think that would be appropriate for the Mars One winners, send them a houseplant for their new home.

Neat question. I can make the attempt. It's cheater cell biology. Cancer is largely a numbers game. A cell has to amass multiple mutations that both provide unlimited proliferation, and knocks out programmed self-destruct. When we're dealing with trillions of cells in a human body, all replicating in a way that is designed to be close - but not perfect - this still takes a long time. First, all cells in an organism come from a single cell. This is fairly elegant as this means that significant errors in replication result in failure of the organism and miscarriage. Minor errors could be accepted though, so some people will be at higher risk for cancer than others. Second, germ cell lines in women freeze replication (in meiosis) at about 12 weeks gestational age. So these cells have undergone significantly less replication than the rest of the 30 year old body. Third, a single spermatozoa, while from older cells as germ cells produce sperm continuously, both needs at least adequate genetics to provide the machinery to make the long swim. The spermatozoa also is much smaller, essentially containing only the genetic material. So the newly formed baby depends on the much younger (relative to number of cell replication and devisions) ova's machinery to replicate. And even if this spermatozoa is somehow inadequate, paired chromosomes mean that most cells have at least two copies of blueprints for everything that needs to be made, and generally, only one copy has to work. Fun stuff, biology.

Somehow I don't think that dramatic descriptions of him cutting off the heads of police are going to help him much.

I would like to hear more about why they are stupid.

Also because when you are trying to pull data, instead of using a handful of codes that encapsulate all the stuff you are trying to study, you have to perhaps draw from a much wider pool. Say you are looking for all ankle fractures - do you now look for ankle fractures, tibial fractures, fibular fractures, injuries caused by giraffes, falls from a height greater than 6.7m but less than 9.5m, etc. There is much potential crossover and yet it may not be fully inclusive of all you want to study. If people who studied in one location tend to code injuries from reptiles, while from another they specify turtles and we see a difference between the two groups, is it because reptiles are different than turtles, or because they were all turtles but one group treats turtle injuries better than the other.

More information is not necessarily better. Case in point: Full body scans - we find more anomalies, but not necessarily more disease as that stone in your gallbladder may never have caused you problems, but now that it is known we are much more likely to want to take it out.

I think this will result in a lower signal to noise ratio.

However, from an epidemiological standpoint, having better information about the causes of health problems will allow better study of cause and effect relationships between wellness and disease, for example. Even if it is a pain to implement (there must be a code for that).

This should intuitively be true. However, making assumptions about a population based solely on someone's ability to find the proper code will not garner the response you are looking for. If I see someone for a twisted ankle (alligator-induced), a common cold, dirt in the eye of extraterrestrial origin, and insulin dependent type 2 diabetes, it is much easier for me to put down the common cold (and perhaps diabetes mellitus, though no specific sub-classification) because I see it all the time and remember it. Most current applications only provide spaces for 3 diagnoses anyhow. ICD 9 gives 780 for general symptoms, and people with much more exacting diagnoses are given this code because it is easy to remember and can be used for almost any situation.

Studies based on retrospective chart reviews often come up with bizarre correlations. Take autism and vaccinations, or acetaminophen and Type I diabetes for example.

I think this has little to do with it. Ultimately we have scientists with a very robust theory, evolution. We currently have a situation of significant selective pressure, as evidenced by the increased number of extinctions that have occurred in recent times. We also have a number of species that we think are at risk for extinction. We are merely trying to determine if our understanding of evolution is sufficient that we can take species which appear to be unable to deal with the selective pressure they are facing and turn them into species that are able. You do this by looking at the whole equation of habitat, predator-prey, population, etc. It is the logical next step - practical application of the theory.

In order to do this, you need to pick species that are doing poorly, but with favourable equations. Meaning that you put your resources into the ones that you think require fairly small changes to turn them around. The ones that have the odds stacked so far against them would be fascinating to succeed with, but would take a long time, and be subject to too many variables to clearly show the benefit resultant from your efforts.

Benefits from this application may eventually be there for those with deep pockets, once the basic science is hashed out. Not that there aren't a lot of bleeding hearts out there to fund the research, but they are most interested in the cute ones, the large ones, and the ones most like themselves (mammals, social structures, etc).

Personifying Nature only is there to create a god out of the mashup of rules that exist out there and create another moral entity, there is no such thing as natural when speaking of disease or predators. Invasive species are simply the ones better adapted to their current environment. This is nothing more than finding a complex machine and trying to figure out how it works.

As an online discussion grows longer, the probability of a comparison involving Nazis or Hitler approaches 1. However, the moment Godwin's Law is mentioned, the probability equals 1.

Nah, it's just the heavy drinkers usually drive alone, and the non-drinkers are driving their family. So when the two collide on the road it's around a 1:5 mortality.

I disagree, While he certainly broke bread, there is no evidence that he did so with the purpose of putting ham, swiss cheese, peppered mayo, lettuce, cucumbers and tomato between the pieces. By the way, I own the copyright of this, and you now are not allowed to make one for yourself without paying me royalties.