The Mann Whitney U test (Wilcoxon Mann Whitney test) part 1/2

แชร์
ฝัง
  • เผยแพร่เมื่อ 3 ธ.ค. 2024

ความคิดเห็น • 10

  • @nikeforo2612
    @nikeforo2612 3 ปีที่แล้ว +3

    Thanks for producing this and all the other videos. I have a lot to catch up....you have been covering very interesting topics. Thanks a lot for spending time in producing this high-quality content. Also, I appreciated the reference to the H0 which is about the probability for the values of one group to score higher than the values of the other....this is often mistaken, and the test is at times considered a test for a difference in median (which is not always true).

    • @tilestats
      @tilestats  3 ปีที่แล้ว

      Thank you! Yes, a H0 that states equal medians can be misleading.

  • @rbc812
    @rbc812 11 หลายเดือนก่อน

    Does it matter "Before" column minus "After" column or the other way round?

    • @tilestats
      @tilestats  10 หลายเดือนก่อน

      No, it should not matter.

  • @brazilfootball
    @brazilfootball 2 ปีที่แล้ว

    Are within group correlations between observations of the same group allowed when using the Mann-Whitney U? Are GLMMs preferable in this case?

    • @tilestats
      @tilestats  2 ปีที่แล้ว +1

      As long as the individuals are independent in the two groups, it is OK to use MW test. Have a look at my LMM video and see if that fits your problem.
      th-cam.com/video/4bGG02Jsjyc/w-d-xo.html

    • @brazilfootball
      @brazilfootball 2 ปีที่แล้ว

      @@tilestats Thank you!

  • @shortandsweet2767
    @shortandsweet2767 ปีที่แล้ว

    Can we calculate P value in Excel for ordinal data samples , if yes please tell the formula ...please

    • @tilestats
      @tilestats  ปีที่แล้ว

      As far as I know, there is no simple way to compute the WMW test in Excel but you can of course follow the steps in this video and compute the p-value as shown at 7:51 by
      =NORM.DIST(-2,02;0;1;TRUE)*2
      or
      =NORM.DIST(-2.02,0,1,TRUE)*2
      depending on your version of Excel.
      Also, if you have ties, you need to use a more complicated formula to calculate the U-statistics, see Wikipedia.

    • @shortandsweet2767
      @shortandsweet2767 ปีที่แล้ว

      @@tilestats thanks