... oh no xkcd

[kaberett]

"Confidence interval": it me.

[Metadata; transcript not available. The comic is a grid of graphs, titled "Curve-fitting methods and the messages they send". Each graph shows the same data, with a different curve. The subgraph entitled "Confidence interval" is captioned "Listen, science is hard but I'm a serious person doing my best."]

Comments

[kaberett]

I'm not sure whether the best bit is that

(i) "Confidence interval" is in fact literally what I'm doing with my current dataset, no, like, right now, or
(ii) that the other publications in the area have broadly taken the approach of the immediately adjacent graph, "piecewise": "I have a theory, and this is the only data I could find."

[cadenzamuse]

:D :D :D

(If I were doing graphs rather than innumerable tables, they would be logistic. But the description is super not applicable, it's just that I have a research question where the dependent variable is yes or no.)

[davidgillon]

"piecewise": "I have a theory, and this is the only data I could find."

... "that didn't blow a hole in it" ;)

Mind, I think piecewise is probably closest to the discontinuous stuff we used to do at work.

[silveradept]

Those graphs were fun to look at, and there were some pretty familiar feelings from many of them.

[sebenikela]

this was the most #relatable xkcd in a WHILE. and that one is also me.

except now i'm in a field where "what actually even is the uncertainty around these projections" is a major subfield of its own. Because APPARENTLY trying to model the energy/landuse/climate/etc systems of the entire world is complicated.