3 ms·

To be frank, this strikes me as concern trolling, assuming this post was made in good faith, perhaps this will enlighten you: https://xkcd.com/1732/ https://xk

by DC-3 8y ago

To be frank, this strikes me as concern trolling, assuming this post was made in good faith, perhaps this will enlighten you: https://xkcd.com/1732/ https://xkcd.com/1732/

mmt 8y ago
> assuming this post was made in good faith That assumption is exhorted by the site guidelines. > https://xkcd.com/1732/ https://xkcd.com/1732/ The problem is that this is merely another "meta report" that the parent specifically found unconvincing. It neither provides hard numbers, nor methodology (which would demonstrate the "hardness", as it were, of those numbers).
- peterashford 8y ago
  The trouble is - that historical data is the data so far as we know. I don't see anyone claiming the historical data is incorrect or fabricated and unless there's a really good reason to doubt it, we should accept it. If you're not convinced by the historical record, then it's more a matter of stubbornness than methodology at that point
  - mmt 8y ago
    > The trouble is - that historical data is the data so far as we know. The trouble is, that the XKCD cartoon is not the data. It is a picture representation of some data. It still falls under the category of "meta" report. This is, of course, as the GP pointed out, "fine", but it does not subsitute for the data itself, nor for the methodology (which is discussed elsewhere in the thread. > I don't see anyone claiming the historical data is incorrect or fabricated I don't personally know enough about the topic to know if that's true, but, given the history of (arguably manufactured) controversy around the topic, nothing would surprise me. Regardless, even as recently as 1,000 years ago, we would not have had reliable, direct temperature measurement across the globe. That means that all but a minute portion of the time period covered by the referenced picture uses synthesized temperature data. Despite being potentially synonymous, I recognize that's not the same as frabricated. However, it does make the methodology that much more important. The data (raw and synthesized) also have other characteristics significant to their eventual interpretation/analysis, besides correctness (or certainty), such as sample frequency. (The concept of overfitted models, for example, was recently discussed on HN). > unless there's a really good reason to doubt it, we should accept it. I posit this stance is incompatible with the scientific method and has contributed to the "reproducibility crisis" in modern science. Although I certainly agree that attempting to reproduce (i.e. gather) raw data should be left as a last resort, I don't feel even it deserves any special treatment. Sadly, the raw data is generally simply not publically available (for science in general, no ideal about climate, specifically). To me, that's the difference between science and religion, with the former, it's actually helpful to "doubt" (i.e. skeptically verify) even if you "believe" (in the conclusion and how it was reached).

by DC-3 8y ago

To be frank, this strikes me as concern trolling, assuming this post was made in good faith, perhaps this will enlighten you: https://xkcd.com/1732/ https://xkcd.com/1732/

mmt 8y ago

> assuming this post was made in good faith That assumption is exhorted by the site guidelines. > https://xkcd.com/1732/ https://xkcd.com/1732/ The problem is that this is merely another "meta report" that the parent specifically found unconvincing. It neither provides hard numbers, nor methodology (which would demonstrate the "hardness", as it were, of those numbers).

peterashford 8y ago
The trouble is - that historical data is the data so far as we know. I don't see anyone claiming the historical data is incorrect or fabricated and unless there's a really good reason to doubt it, we should accept it. If you're not convinced by the historical record, then it's more a matter of stubbornness than methodology at that point
- mmt 8y ago
  > The trouble is - that historical data is the data so far as we know. The trouble is, that the XKCD cartoon is not the data. It is a picture representation of some data. It still falls under the category of "meta" report. This is, of course, as the GP pointed out, "fine", but it does not subsitute for the data itself, nor for the methodology (which is discussed elsewhere in the thread. > I don't see anyone claiming the historical data is incorrect or fabricated I don't personally know enough about the topic to know if that's true, but, given the history of (arguably manufactured) controversy around the topic, nothing would surprise me. Regardless, even as recently as 1,000 years ago, we would not have had reliable, direct temperature measurement across the globe. That means that all but a minute portion of the time period covered by the referenced picture uses synthesized temperature data. Despite being potentially synonymous, I recognize that's not the same as frabricated. However, it does make the methodology that much more important. The data (raw and synthesized) also have other characteristics significant to their eventual interpretation/analysis, besides correctness (or certainty), such as sample frequency. (The concept of overfitted models, for example, was recently discussed on HN). > unless there's a really good reason to doubt it, we should accept it. I posit this stance is incompatible with the scientific method and has contributed to the "reproducibility crisis" in modern science. Although I certainly agree that attempting to reproduce (i.e. gather) raw data should be left as a last resort, I don't feel even it deserves any special treatment. Sadly, the raw data is generally simply not publically available (for science in general, no ideal about climate, specifically). To me, that's the difference between science and religion, with the former, it's actually helpful to "doubt" (i.e. skeptically verify) even if you "believe" (in the conclusion and how it was reached).