seriousstats.bsky.social
Mostly posting about statistics and Psychology.
55 posts
698 followers
1,177 following
Prolific Poster
Conversation Starter
comment in response to
post
They are going to struggle to keep or recruit top researchers if they can't publish in the journals that (for better or worse) define your career. Of course that might be the plan ...
comment in response to
post
Not quite the same but reminds me of the classic linguistics joke/story about double positives: www.ling.upenn.edu/~beatrice/hu...
comment in response to
post
* "understand" ... not "understanding" ...
comment in response to
post
Of course if the only measures have terrible reliability (e.g, .5 or lower) you are probably better off using limited resource to get a better measure first than trying to increase n to compensate for the poor reliability.
comment in response to
post
I actually think it's a bit of a red herring. You want to use measure with best available reliability and validity (or your best bet as it - because its arguably a property of the sample plus measure not the measure alone). Though maybe I'm the only one who thinks that ...
comment in response to
post
"In the early stages of research . . . one saves time and energy by working with instruments that have only modest reliability, for which purpose reliabilities of .70 or higher will suffice…"
(Nunnally, 1978, p.245).
comment in response to
post
Good to see them mention that .70 is not what Nunally recommended. For context here is the longer Nunally quote:
comment in response to
post
Isn't this always. No. At best our models capture the key features of the data structure are relevant to produce a good enough model of the data to answer the research Qs. The overlap with forensic data analyses is I think more in presenting and describing the data structure accurately and honestly.
comment in response to
post
That is not doxxing. It's reporting newsworthy information on someone who, in taking a "government efficiency" job with Elon Musk, placed themselves in the public eye. The public has a right to know who the people working for Musk are and what they stand for.
comment in response to
post
Giving access to SSNs, other data on government employees to non-employees like Musk is a violation of 5 USC 552a, and carries with it a penalty of $1000 per person whose data was accessed. We are building a case. Govt employees who want to participate should contact me directly
comment in response to
post
I'm not so sure ... lots of pre-registrations are for funding bids. It's be interesting to see - I'd guess it would be a fair chunk.
comment in response to
post
Psychology Starter Packs
Mega-thread of 68 psychology-related starter packs!
Click on this post and scroll up! ☝️
#AcademicSky #PhDSky 🧪