Thread by @FiPanther

I have so many spicy astronomy takes that I could probably do an entire thread.

For every like this tweet gets, I will add a spicy astronomy take™️

1. Your subgrid model needs to be motivated by the underlying physics first, and tuned to observations second

2. All observations are inherently biased, and thinking your special way of observing the universe isn't biased just demonstrates that it is biased

3. There is a lot more to Bayesian statistics than plugging and chugging through an equation or python package

4. If your error bars are that big, you can't make dramatic conclusions. People need to be more cautious

5. 'One solution to the astrophysical problem to rule them all' is satisfying and gets nature papers, but probably not realistic

6. Reducing your computational method to 'we did a simulation using method' and not giving a detailed description is ruining the lives of PhD students everywhere

7. Crapping all over 'tilde astronomy' because someone ruled something out without chewing up 10 kSU of compute time is really passé

8. Theorists need to try out making observations, reducing and analysing them before critiquing observational studies.

9. Observers need to start listening to theorists and statisticians before drawing dramatic conclusions from their observations

10. If the underlying process isn't Gaussian, STOP TALKING ABOUT SIGMA VALUES. Go talk to a statistician

11. Being nice and listening to the technical staff at the telescops is really the least you can do.

12. Don't accept invitations to comment in the media on something you didn't do any work on and make it look like you were directly involved in the project. Comment on the work, or better yet, ask the media person to contact the lead authors or people directly involved.

13. Astronomers aren't experts on every kind of science and we need to stop pretending to be. Media want to talk to you about mortuary science? Recommend a mortician for them to contact. Biology? Biologist. Climate science? Climate scientist.

14. Spend less time talking doom and gloom about the '2% become professors' statistic and work on helping mentees be best prepared for academic job applications, and help them divest interests in case it doesn't work out as they hope.

15. You should be going to colloquia outside your field of expertise. Just because you don't work on what the speaker is talking about doesn't mean you won't learn something valuable

16. Publishing papers as an undergrad is becoming a prerequisite for getting into grad school and it needs to not be. Students who excel at research may not be afforded UG research opportunities

17. Machine learning isn't a black box and using it as one is really fraught with problems. Understand the algorithm before using it.

18. GPUs are massively underutilized in astro and GPU programming needs to be taught in classes. Your GPU isn't just for making fortnite look cool on your computer

19. It isn't 'a code'. It's a 'program'. You didn't write a 'code to simulate x', you wrote a 'computer program to simulate x'

20. Stop putting telescopes on land that belongs to and is of spiritual importance to indigenous peoples.

21. Tell indigenous stories about the night sky when doing outreach as well as the usual Greek myths.

22. If an astronomy job advert is tailored for one specific person, it shouldn't be publically advertised and you should be honest with people expressing interest in the job. There needs to be some kind of rule change around this.

23. PhD students should be paid a fair, at minimum living wage, and they should not have to teach or work to support themselves through what is effectively a full time job.

24. Paid sick leave and bereavement leave during a PhD program should be a thing everywhere in the world.

25. If a student or postdoc has shown the initiative to organize and run group meetings etc, senior scientists should be resisting to trying to muscle in and take over running said meetings

26. Do we really need to continue using proprietary software to do astronomy?

27. Does data really have to be proprietary? If you don't release your raw spectra, how can I be sure your results are actually reproducible?

28. The spiciest take yet: there is a lot more to astronomy than galaxy evolution.

Yes. I said it.

29. Having a high instrumental background isn't a problem as long as you understand the background well and have good statistical techniques to remove it.

30. If you don't have a special dietary requirement, STOP EATING THE FOOD FOR PEOPLE WITH SPECIAL DIETARY REQUIREMENTS.

31. Telling students 'I didn't have an amazing GPA and I still got into grad school' can seem kind of encouraging, but we need to be aware that a) times have changed and b) some students depend on scholarships which are awarded based on maintaining GPA

32. See above but for astronomy postdocs and 'well I only had 1 paper and I got a postdoc'. Competition for level A positions these days means competing with people with 5+ years of postdoc experience.

33. Supervisors should only take on a limited number (less than 5-6) students at all levels to ensure adequate supervision time for all

34. Highly specific, but while astronomy departments seem to be relaxed places in general, but this shouldn't mean you can bring your dog to work. People are afraid or nervous around dogs regardless how much of a 'good doggo' yours is.

35. Students need to understand the basic physics before going into astronomy. If you want to study spectral line formation you should probably take some atomic physics classes.

36. The only way to get better at academic writing is to read lots of papers, both in your field and in adjacent fields. Just because the paper isn't on exactly what your work is doesn't mean reading it is a waste of time

37. People who take breaks to attend social events like morning tea publish at a higher rate than those that don't (I have data on this one)

38. If you are going to take on leadership of a large collaboration, lab or project, taking classes in project management will save a lot of heartache for you and the people you are managing

39. Stop crapping all over scientists that go and work in the defense industry. Stop crapping all over the military. Sure you have your ethical reasons to 'hate' the military, but these are the people who are protecting you in a time of war.

I see this so much from astronomers

39 addendum: during previous wars, I am under the impression ssipn that physicists were exempt from being drafted. That's a hell of a privilege.

40. Cube farm style shared office spaces are a horrible idea.

41. 'Mental health' isn't limited to supporting people with anxiety and depression. Academics have higher incidences of conditions such as bipolar disorder than the general population. Educate yourselves about more than just a catch-all term

42. You can disagree with someone's scientific ideas and still respect them as a person.

Ok no more spicy takes for now. I am going to write my thesis and let the wrath of Twitter swirl around this thread.

To clarify this, I have no issue with bringing pets to work (or children) if they are in an office and meeting the animal isn't forced on people while they're eating, walking around etc. Phobias of certain animals are real.

43. I think I'm just pretending to enjoy plasma physics at this point because... Fake it till you make it?

44. The simpler the elemental particle propagating through space, the more complex it's interactions are, at least in my experience

45. Not really a spicy take but a fact. If we hadn't blown up nuclear weapons, the Galactic positron annihilation signal would have been discovered 10 years earlier rather than being attributed to a mysterious Ru109 decay gamma ray line

46. Gamma ray astronomy and cosmic ray astrophysics is massively underappreciated

47. I believe that if we discover dark matter, we're gonna discover it by indirect detection in an astrophysical context. I would love to be proven wrong

48. Supervisors should be encouraging their students to attend journal clubs, not dismissing them as a waste of time. Same goes for seminars etc.

49. International travel to conferences is great but not accessible to everyone, especially for scientists who may not be able to travel through the middle east for various reasons

50. I'd love to see more people my age/career stage being invited to conferences in different disciplines. Is it only gamma ray astronomy that is excited to give young scientists opportunities like keynote invited talks?

51. There is no excuse for over running your allocated talk slot at a conference. I don't care if it's a plenary or a contribution. Nobody is so important they don't have to keep to time

52. If you and someone else are giving similar conference talks, for crying out loud contact each other to make sure you don't just present the same identical material and can complement each other. All it takes is an email convo

53. I believe all conferences should have a 'do not disturb the scientists' chill out zone where people can go to escape constant conversation and stimulation.

54. Fortran is an underrated programming language for astronomy

55. JSON is a gross and overrated format for astronomy

56. Making data public is one thing, obfuscating access to it is quite another (looking at you, SDSS)

57. Just because you commented your code doesn't make it automatically distributable, unless you list all the library dependencies and fudge factors you use. I could list so many astro software packages that are barely useable because of terrible documentation

58. The next few hot takes brought to you by @DRG_physics:

It isn't MOND, no matter how much you want it to be

59. Astronomy is still astronomy. Just because you like your part of the spectrum best doesn't make it superior.

60. Twitter is a really good resource for ECRs and some senior academics should stop bashing it.

61. You don't have to fit some cookie cutter expectations to be an astronomer, just do you. By the same token, media needs to recognize that there are a lot of ways to be astronomers and stop just showing the cookie cutter ideal.

Thanks to @DRG_physics for these takes

62. Back to my takes:

Astronomy isn't better than any other science. You aren't on some higher plane of existence like the expanding brain meme. It's a science just like any other and it is also immensely flawed

63. Astronomy wasn't 'invented' by Galileo in the 1600s. Indigenous people have been doing astronomy for 10s of thousands of years and this needs to be acknowledged way more in text books.

More with help from @DRG_physics reminding me:

Weekend conferences, and conferences that run for more than 3-4 days, or back to back conferences are not family friendly. Also just because it's in school holidays doesn't make it family friendly

65. Disregarding or dismissing someone as a researcher until they have something they can offer you or a skill you can use them for is rude af and can we stop doing this

66. Nobody understands your 'simple' equations in your conference talk. Take them out.

67. No field of astronomy is inherently better than any other, or more valuable. Stop dismissing others work because it doesn't fit in your fake hierarchy

68. While you're whining that your already built telescope had been delayed, others are desperately trying to convince the powers that be that their instrument, the only in its wavelength range, should be funded at all.

69. The 'relevance to society' or whatever part of funding proposals is actually kind of important. It shows you can communicate your work and excite a general audience about what you are doing, and that you can think more broadly than your own h-index

70. It's all very well to point out that astronomy and academia are biased toward the success of certain racial and gender groups, but if you are in a position of power and not actively doing something to change this, it's lip service

71. Painting something pink or associating science with female gender stereotypes is lazy activism. There are a million ways to promote science to diverse gender identities without making a 'science cookbook for girls' or something, which is actually pretty alienating

72. It would be cool if 'women in STEM' initiatives could be more inclusive to gender diverse people tbh

73. If you publish something which is similar to an earlier work, it's generally nice to cite the earlier work and show that yours acknowledges that work

74. Picking a graduate school based solely on its ranking is not a great idea. Picking a supervisor based on their fame is not a great idea. You're better off picking a supervisor you get along with.

75. Corollary to 74: it's pretty unkind to say to someone 'oh, your school isn't very highly ranked' or 'I would never want to go to school there' or 'why would you work with x? Nobody knows who they are'

76. Corollary to 75: however to some extent having a good, solid reference from someone well known who knows your work well is usually a good thing if applying for postdocs, so choose your collaborators wisely I guess?

77. If you're worried someone is going to oppose some work you're doing, be proactive and approach them to discuss your work, especially if it's a bit controversial

78. 🌶🌶🌶 incoming:

Papers belong on the arxiv when they've been submitted.

If the paper may get rejected (unless it's been sent to a 'big journal' like nature or science, where rejections are super common), it isn't ready to be submitted

Ok, no more. I'm sorry guys. My notifications are a dumpster fire and I can't reply to everyone rn, and I'm gonna go chill and have a Christmas eve. Thank you for distracting me from my thesis. I love you all!