ASA statement on P-values and statistical significance: Development and impact

แชร์
ฝัง
  • เผยแพร่เมื่อ 3 พ.ย. 2024

ความคิดเห็น • 10

  • @Felicidade101
    @Felicidade101 6 ปีที่แล้ว +1

    who is that man that speaks @1:04:03? and did he publish the different problems remark?

  • @slrellison
    @slrellison 7 ปีที่แล้ว +2

    One important thing the ASA statement does not comment on is the role of the p-value in keeping bad papers _out_ of the literature. Imperfect it may be, but if that simple threshold were not out there, there's be a heck of a lot more irreproducible studies in the literature.

    • @ankanbiswas4321
      @ankanbiswas4321 6 ปีที่แล้ว

      Agreed. But it filters to a certain extent. There are so many junk papers out there with irreproducible results but with p value less than 0.05.

    • @chadiao791
      @chadiao791 6 ปีที่แล้ว +3

      slrellison it seemed like you favored publication bias using p-value as a screening tool

    • @TheBrocktonBlockbust
      @TheBrocktonBlockbust 5 ปีที่แล้ว +1

      @@chadiao791 I do agree. The problem of nowadays publication bias, or at least a part of it, is due to the fact that people and journals tried to apply a dichotomus standard to papers between Statistically Significant and not; its the same as saying "good and bad".

    • @alejandrarodriguezsanchez6667
      @alejandrarodriguezsanchez6667 5 ปีที่แล้ว +2

      no! that is precisely the problem! good done research should not have anything to do with pvalues!!!!

  • @wenmichael6596
    @wenmichael6596 5 ปีที่แล้ว +1

    Can OUP republish those Fisher’s classic three books ? So we can extract his original ideas,