Ruben C. Arslan

@rubenarslan

Bayescurious evidence enthusiast . Topics: evolution, ovulation, mutation, intelligence, personality, sexuality, R, open science & source tools.

Berlin, Germany
Đã tham gia tháng 4 năm 2013

Tweet

Bạn đã chặn @rubenarslan

Bạn có chắc muốn xem những Tweet này không? Xem Tweet sẽ không bỏ chặn @rubenarslan

  1. Tweet đã ghim
    31 thg 5, 2021

    Now out in Evolution and Human Behavior We clarified the scope and range, i.e., we didn't look at long-term attraction, nor did we include men with IQs below ~80. For me, this ends my interest in "intelligence as a fitness indicator", but YMMV.

    Hiện chuỗi hội thoại này
    Hoàn tác
  2. đã Tweet lại
    24 giờ trước

    And "I am also excited to announce that JPSP: PPID will now consider registered reports for any submission type (see for useful information about the registered-report format)." Thanks for your leadership !

    Hiện chuỗi hội thoại này
    Hoàn tác
  3. 19 thg 1

    It seems iOS recently changed their German autocorrect algorithm. All iPhone messages I get suddenly call me Rubén. I can get into it! I just hope they don't switch back all of the sudden once I've really leaned into my new persona.

    Hoàn tác
  4. đã Tweet lại
    17 thg 1

    The contribution of dominance genetic effects is minimal compared to additive genetic effects, concludes this new study on >1000 complex traits in ~360k participants. -

    Hoàn tác
  5. đã Tweet lại
    14 thg 1

    Check out our new article in comparing LC-MS assessments of progesterone, cortisol and testosterone in saliva and hair across two menstrual cycles in 155 women. Led by and with .

    Hoàn tác
  6. đã Tweet lại
    13 thg 1

    This example is extreme, but the underlying practice is common. You end up with scales that just ask the same over and over again — it really shows how in our field, we care more about “optimizing” stats than about our participants.

    Hiện chuỗi hội thoại này
    Hoàn tác
  7. đã Tweet lại
    13 thg 1

    Saw a scale with 66 items to measure something* that can probably be assessed with 1, 2, maybe 3 items if you really care about it. That’s what we get for acting as if Cronbach’s alpha was the most important thing about a scale. (*It’s the Sex Doll Ownership Attitudes Scale).

    Hiện chuỗi hội thoại này
    Hoàn tác
  8. đã Tweet lại
    13 thg 1

    Phenomenal paper just published in Science persuasively demonstrating that EBV causes most if not all multiple sclerosis. Phenomenal epidemiological design and technical achievement. 👏

    Hiện chuỗi hội thoại này
    Hoàn tác
  9. 13 thg 1

    How should we respond to people who misreport something on a journal disclosure form? E.g. ticked box saying they shared code/data or disclosed why not but actually did no such thing. How do you respond?

    Hoàn tác
  10. đã Tweet lại
    12 thg 1

    On CRAN very very soon !

    Hiện chuỗi hội thoại này
    Hoàn tác
  11. 12 thg 1

    Deadline to apply end of January. I worked for Michael ten years ago and quite enjoyed it, DM me know if you want to know more.

    Hoàn tác
  12. đã Tweet lại
    11 thg 1

    Wir suchen eine Person für eine PostDoc-Stelle mit Schwerpunkt Forschung im Bereich Psychologische Diagnostik, Psychologische Methoden oder Persönlichkeitspsychologie an der HMU Potsdam. Bitte RT

    Hoàn tác
  13. đã Tweet lại
    11 thg 1

    Even in retrospect, the story of how we discovered the cure for scurvy is pretty complicated. But we worry that it would have looked even messier from the inside.

    Hiện chuỗi hội thoại này
    Hoàn tác
  14. 11 thg 1

    "a linear combination of the average height of parents and of the offspring’s PGS yields an unprecedented accuracy of 54% (SE 3.2%)." using genetic data alone you could get a perpetrator's height down to ±5.4cm (assuming 7cm SD, known sex)

    Hoàn tác
  15. đã Tweet lại
    10 thg 1

    Autism (+) and ADHD (-) genes seem to have modest causal effects for Educational Attainment independently of IQ. Mendelian randomization.

    Hoàn tác
  16. đã Tweet lại
    10 thg 1

    2) We want researchers to be transparent & make testable claims, but we prefer claims that are not wrong. That's a strong incentive for making claims that are difficult to test, either b/c they are too vague or because the test requires knowledge/equipment that isn't public. 4/

    Hiện chuỗi hội thoại này
    Hoàn tác
  17. đã Tweet lại
    10 thg 1

    Why I think CODA-O would be hard to get off the ground and maybe impossible to maintain: 1) We don't want researchers to overclaim, but we prefer more impressive claims. That's a very strong incentive for overclaiming, and it's only held in check via methodological standards. 3/

    Hiện chuỗi hội thoại này
    Hoàn tác
  18. đã Tweet lại
    7 thg 1

    The PGS portability issue is as strong as reported before, even when training and testing in the same cohort.. Relative variance explained by PGS (trained in UK) is 86% for S. Europe, 65% for S. Asia, 49% for E. Asia, and only 18% for W. Africa (on avg, compared to UK). 4/6

    Hiện chuỗi hội thoại này
    Hoàn tác
  19. đã Tweet lại
    7 thg 1

    Except for a few outliers, most phenotypes follow the same trend of portability. On average, portability is decreasing proportional to the genetic distance between populations. 5/6

    Hiện chuỗi hội thoại này
    Hoàn tác
  20. 7 thg 1

    "His mind is so often blown that it’s little wonder it has such difficulty in paying attention." Oof.

    Hoàn tác
  21. đã Tweet lại
    6 thg 1

    Delighted to say I’ll soon be starting a 3-year postdoc working on this exciting meta science project with , & ! (the good kind of meta, not the zucc)

    Hoàn tác

Tải trang có vẻ sẽ mất một lúc.

Twitter có thể đang bị quá tải hoặc tạm thời trục trặc. Thử lại hoặc truy cập Trạng thái Twitter để biết thêm thông tin.

    Bạn cũng có thể thích

    ·