Thanks for researching and writing this up! Weâve been discussing the topic a lot at CEA/âGiving What We Can over the last few days. I think this points to the importance of flagging publication dates (as GiveWell does, indicating that the research on a certain page was current as of a given date but isnât necessarily accurate anymore). Fact-checking, updating, or just information flagging as older and possibly inaccurate was on our to-do list for materials on the Giving What We Can site, which go back as much as 10 years and sometimes no longer represent our best understanding. I now think it needs to be higher priority than I did.
For individuals rather than organizations, Iâm unsure about the best way to handle things like this, which will surely come up again. If someone publishes a paper or blog post, how often are they obliged to update it with corrected figures? Iâm thinking of a popular post which used PSIâs figure of around $800 to save a childâs life. In 2010 when it was written that seemed like a reasonable estimate, but it doesnât now. Is the author responsible for updating the figure everywhere the post was published and re-published? (Thatâs a strong disincentive for ever writing anything that includes a cost-effectiveness estimate, since theyâre always changing.) Does everyone who quoted it or referred to it need to go back each year and include a new estimate? My guess is itâs good practice, particularly when we notice people creating new material that cites old figures, to give them a friendly note with a link to newer sources, with the understanding that this stuff is genuinely confusing and hard to stay on top of.
Itâs obviously impossible to enforce everyone to update figures all the time. If there is an old publication date, everyone probably understands that it could be outdated. I just think that the date should be always featured prominently. E.g. in this page it could be better. I think that flagging pages the way GiveWell does is a great idea. But featured pages that have no date should probably be checked or updated quite often. I mean pages like âtop charitiesâ, âwhat we can achieveâ and âmyths about aidâ in GWWCâs case.
Thanks for researching and writing this up! Weâve been discussing the topic a lot at CEA/âGiving What We Can over the last few days. I think this points to the importance of flagging publication dates (as GiveWell does, indicating that the research on a certain page was current as of a given date but isnât necessarily accurate anymore). Fact-checking, updating, or just information flagging as older and possibly inaccurate was on our to-do list for materials on the Giving What We Can site, which go back as much as 10 years and sometimes no longer represent our best understanding. I now think it needs to be higher priority than I did.
For individuals rather than organizations, Iâm unsure about the best way to handle things like this, which will surely come up again. If someone publishes a paper or blog post, how often are they obliged to update it with corrected figures? Iâm thinking of a popular post which used PSIâs figure of around $800 to save a childâs life. In 2010 when it was written that seemed like a reasonable estimate, but it doesnât now. Is the author responsible for updating the figure everywhere the post was published and re-published? (Thatâs a strong disincentive for ever writing anything that includes a cost-effectiveness estimate, since theyâre always changing.) Does everyone who quoted it or referred to it need to go back each year and include a new estimate? My guess is itâs good practice, particularly when we notice people creating new material that cites old figures, to give them a friendly note with a link to newer sources, with the understanding that this stuff is genuinely confusing and hard to stay on top of.
Itâs obviously impossible to enforce everyone to update figures all the time. If there is an old publication date, everyone probably understands that it could be outdated. I just think that the date should be always featured prominently. E.g. in this page it could be better. I think that flagging pages the way GiveWell does is a great idea. But featured pages that have no date should probably be checked or updated quite often. I mean pages like âtop charitiesâ, âwhat we can achieveâ and âmyths about aidâ in GWWCâs case.