4 ms·

It's not easy to find another situation where no other confounding effects were present. As explained in the article, the scholar brings good arguments as to wh

It's not easy to find another situation where no other confounding effects were present. As explained in the article, the scholar brings good arguments as to why this is as close as it gets to a randomised trial.

chiefalchemist 8y ago
I didn't say it was easy :) But that's still no excuse to say B leads to C when in fact, it (probably) does not. Science isn't supposed to be easy. Truth isn't supposed to be easy. If it's easy, then there's a fair chance it's not either one of these.
- ineedasername 8y ago
  Why would you say it probably does not? Aren't you falling into the same lack of evidence trap you claim is present here? As an aside, this conclusion is bolstered by research into factors influencing student outcomes in modern time: Even controlling for socioeconomic factors, students from more educated families and peer groups tend to complete higher levels of education, and thus perpetuate that cycle with their own children later in life. That's the same cycle being observed in this article. There may be other confounding factors that were the cause here, so further examples would be enlightening, but as it is this is a result that fits very well within the existing body of knowledge regarding factors that influence educational outcomes.
  - chiefalchemist 8y ago
    Not is the baseline. I'm not drawing any conclusions until there's proof otherwise. Note: (probably) was in parens in a wink wink editorial sorta way. That is, I didn't want to say it as fact. But hint, given the lack of absolutely evidence, otherwise, it's a pretty safe presumption.
    - ineedasername 8y ago
      Wow, you say you're not drawing a conclusion and then "wink wink" go on to do exactly that, and in contradiction to the article without proof or even a rationale for why you seem disinclined to believe in the stated statistical significance on display here. Statistics are never "absolute" in the way you seem to be demanding, they're a matter of probability and likihood of chance outcomes. The outcomes cited here are well within established bounds of statistical significance and as well they comport with the existing body of research on educational outcomes.
- diggernet 8y ago
  I agree with your observation that in this case A was ultimately the cause of C. But as shinryuu noted above, other missions did not focus on B and did not achieve C. Because of that, it is reasonable to conclude that something else leading to B would also lead to C. So while A was ultimately the cause of C in this case, B is the cause which matters most if you want to reproduce C, because it means you can reach C with or without A. Your statement now that B probably does not lead to C puzzles me. If that is true, then the A -> B -> C chain is broken, and your original statement that A is the cause of C (based on that chain) is invalidated. So I don't know where you are going with this.
- alehul 8y ago
  Occam's razor would argue that truth should be rather easy. I think you have a valid point, but that the correlation between B and C is still very likely, even if A is a factor in the existence of B. For instance, if I have a particularly great lunch one day because a new guy is working the grill at a restaurant, I would claim that the new guy working the grill is the reason for my great lunch. There are, of course, many factors which go into him being there, such as the manager who chose to hire him, but it is nonetheless truthfully and directly because of him that I've had a great lunch. Similarly, a mission is relevant and increases the likelihood of an education; it doesn't mean that education is not the most direct correlation, however. Additionally, somebody pointed out how similar effects were not found when missions focused on health and poverty, for instance.

chiefalchemist 8y ago

I didn't say it was easy :) But that's still no excuse to say B leads to C when in fact, it (probably) does not. Science isn't supposed to be easy. Truth isn't supposed to be easy. If it's easy, then there's a fair chance it's not either one of these.

ineedasername 8y ago
Why would you say it probably does not? Aren't you falling into the same lack of evidence trap you claim is present here? As an aside, this conclusion is bolstered by research into factors influencing student outcomes in modern time: Even controlling for socioeconomic factors, students from more educated families and peer groups tend to complete higher levels of education, and thus perpetuate that cycle with their own children later in life. That's the same cycle being observed in this article. There may be other confounding factors that were the cause here, so further examples would be enlightening, but as it is this is a result that fits very well within the existing body of knowledge regarding factors that influence educational outcomes.
- chiefalchemist 8y ago
  Not is the baseline. I'm not drawing any conclusions until there's proof otherwise. Note: (probably) was in parens in a wink wink editorial sorta way. That is, I didn't want to say it as fact. But hint, given the lack of absolutely evidence, otherwise, it's a pretty safe presumption.
  - ineedasername 8y ago
    Wow, you say you're not drawing a conclusion and then "wink wink" go on to do exactly that, and in contradiction to the article without proof or even a rationale for why you seem disinclined to believe in the stated statistical significance on display here. Statistics are never "absolute" in the way you seem to be demanding, they're a matter of probability and likihood of chance outcomes. The outcomes cited here are well within established bounds of statistical significance and as well they comport with the existing body of research on educational outcomes.
diggernet 8y ago
I agree with your observation that in this case A was ultimately the cause of C. But as shinryuu noted above, other missions did not focus on B and did not achieve C. Because of that, it is reasonable to conclude that something else leading to B would also lead to C. So while A was ultimately the cause of C in this case, B is the cause which matters most if you want to reproduce C, because it means you can reach C with or without A. Your statement now that B probably does not lead to C puzzles me. If that is true, then the A -> B -> C chain is broken, and your original statement that A is the cause of C (based on that chain) is invalidated. So I don't know where you are going with this.
alehul 8y ago
Occam's razor would argue that truth should be rather easy. I think you have a valid point, but that the correlation between B and C is still very likely, even if A is a factor in the existence of B. For instance, if I have a particularly great lunch one day because a new guy is working the grill at a restaurant, I would claim that the new guy working the grill is the reason for my great lunch. There are, of course, many factors which go into him being there, such as the manager who chose to hire him, but it is nonetheless truthfully and directly because of him that I've had a great lunch. Similarly, a mission is relevant and increases the likelihood of an education; it doesn't mean that education is not the most direct correlation, however. Additionally, somebody pointed out how similar effects were not found when missions focused on health and poverty, for instance.