@PhDemetri I gave my son a 36" novelty measuring tape. Turns out it's a hammer, can measure tables up to 3 years long, and even works as a retractable ninja whip.
@MelissaGWolf I've always thought that the nomological network is the idealized form. It's what we want when we start... Then, when we realize it's not possible, we start redefining our goals and acting like the NN was never all that great of an idea anyways and we never really wanted it. 🤔🤔
@PhDemetri I’ve followed you mostly for puns, dog pictures, and links to cross validated responses. Big fan of this style and will definitely be checking in.
I’m exactly the sort of person you’re talking about as a journeyman statistician - Also a big fan of that label.
@IsabellaGhement Key note from the boxplot.stats docs: "The two ‘hinges’ are versions of the first and third quartile, i.e., close to quantile(x, c(1,3)/4). The hinges equal the quartiles for odd n (where n <- length(x)) and differ for even n."
I'd say ggplot matches the 'expected' behaviour
@PhDemetri Type S and M error rates are such a cool and under-taught concept.
I got to apply them during my thesis studies and would talk to everyone I could about it. They give a whole new level of understanding to what’s actually going on with freq. stats, imo
@seanpmackinnon A recently completed thesis has some relevant discussion of this issue: https://t.co/sqZzLT9Lrf
The author is a colleague, not on Twitter
@JennyBryan@rstudio Yup. It creates a dissonance by merging/conflicting different senses of the word: "I posit x" has the senses of "I hypothesize x" and "I propose/put forward the idea that x". Stringing together forces rejecting of the former to resolve the juxtaposition: "I propose hypotheses"
@PhDemetri My only solution that makes the anxiety go away: Assume the entire day is a write-off, leave as soon as done packing, and commit to hanging out in the airport directly in front of the gate until it's time to go. Working remote is optional, depending on stress.
@BD_Zumbo @mededdoc This is the way to escape from the despair of “validation never ends”. It may not end, but it can be deemed sufficient for a specific, well defined context/ecosystem of measurement.
Same as a court of law, or any scenario where evidence is used to inform inference.
@BD_Zumbo @mededdoc Reframing this (Accurately, I think?): The answer to whether or not a partial story is acceptable comes down to why & how one is conducting validation (a theory of validity) and why & how the measure is being used. What evidence is required to support the inference being drawn?
@nntaleb@pkedrosky Many activities achieve the same sensation, whatever the reason, but they’re not broadly classified as “exercise”, they’re “sports”. Bouldering, ski/snowboarding, distance swimming, mountain biking. I don’t see how the diff is anything but socially constructed, like @pkedrosky