Poisson Regression Part I | Statistics for Applied Epidemiology | Tutorial 9

Поділитися
Вставка
  • Опубліковано 3 гру 2024

КОМЕНТАРІ • 39

  • @cyg7655
    @cyg7655 4 роки тому +2

    Clear & straightforward. Thank you

  • @caseyli5580
    @caseyli5580 4 роки тому +3

    This was a great video - thank you!
    Only nit is that the brackets at 14:29 are a little misleading - to echo the comment below, you may want to clarify the following:
    ln(E(y)) - ln(t_i) = ln(E(y) / t_i) = ln(y/10000) = -3.29665

  • @kelvinsonmwangi8359
    @kelvinsonmwangi8359 5 років тому +5

    Good explanation. What is the name of the data set used in the presentation?

  • @anapaulaserafinipoetasilva1044
    @anapaulaserafinipoetasilva1044 5 років тому +4

    amazing class!

  • @dataman6744
    @dataman6744 3 роки тому

    Thanks, great explanation

  • @HyacintheTandoro
    @HyacintheTandoro Рік тому

    Thanks for the tutorial

  • @katherinesuter9597
    @katherinesuter9597 2 роки тому +1

    Thank you for the tutorial - very clear and concise. How do you get access to the R code and data files?

  • @abcpsc
    @abcpsc 3 роки тому +1

    12:20 how to get that confidence interval?

  • @bestkebechannel1860
    @bestkebechannel1860 2 роки тому

    very clear, thank you

  • @enricvilar7232
    @enricvilar7232 4 роки тому

    Phenomenally clear. thanks

  • @sjoerdglaser2794
    @sjoerdglaser2794 5 років тому +1

    Very clear tutorial! I understand what poisson regression is now
    Quick question: By doing all those different tests in the end, isn't that a bit fishing? Just doing all kind of different tests and in the end report the one that show the most interesting results?

    • @marinstatlectures
      @marinstatlectures  5 років тому +1

      im a bit unclear on exactly what you are referring to, but here are some general comments that may help. first, you should definitely avoid "fishing" and just randomly poking around and reporting the most significant result. but in building a model and examining a research question, you often want to hypothesize some relationship, and then test to see if it is there statistically. you may also hypothesize an interaction/effect modification, and then wish to test statistically is it is significant (as well as biologically plausible).
      so, it is ok to be doing a series of tests, etc, as long as they are considerations you have made ahead of time, and you are checking to see if they are there statistically...but it is not a good practice to just "test everything" and then report on that...there should be some conceptual basis for what you are doing.
      hope that helps...

  • @kevinescandon
    @kevinescandon 5 років тому +2

    I loved this. Thanks!

    • @marinstatlectures
      @marinstatlectures  5 років тому

      You are welcome Kevin! glad you found the video useful

  • @lauradallars8201
    @lauradallars8201 4 роки тому +2

    Does anyone know if the R code is provided for this series? I really appreciate the help!

  • @qianwenliu7127
    @qianwenliu7127 4 роки тому +11

    Amazing tutorial! But I think there is a little mistake in the formula for the connect rates to counts part. The formula should be ln(E(yi))-ln(ti)=ln(E(yi)/ti) instead of ln(E(yi))-ln(ti)=ln(E(yi))/ln(ti)

    • @programmingwithsadeeq2730
      @programmingwithsadeeq2730 2 роки тому +1

      Hi Qianwen, actually this is not a mistake. And both of you are correct. You can write it either.

  • @pedrolavinas5568
    @pedrolavinas5568 4 роки тому +2

    can we have access to the database used?

  • @ishanbuddhika4317
    @ishanbuddhika4317 4 роки тому +1

    Explanation is AMAZING! Thank you very much! :)

    • @madduxkade1318
      @madduxkade1318 3 роки тому

      I dont mean to be so offtopic but does someone know a method to log back into an Instagram account??
      I stupidly lost my password. I appreciate any assistance you can give me!

    • @deancruz8296
      @deancruz8296 3 роки тому

      @Maddux Kade instablaster :)

    • @madduxkade1318
      @madduxkade1318 3 роки тому

      @Dean Cruz Thanks for your reply. I got to the site thru google and Im trying it out now.
      Seems to take quite some time so I will get back to you later with my results.

    • @madduxkade1318
      @madduxkade1318 3 роки тому

      @Dean Cruz It worked and I actually got access to my account again. I am so happy!
      Thanks so much you saved my account :D

    • @deancruz8296
      @deancruz8296 3 роки тому

      @Maddux Kade you are welcome =)

  • @snigdho642
    @snigdho642 3 роки тому

    thank you for this video!

  • @youtbeuser32
    @youtbeuser32 3 роки тому +1

    where Can I get the data analysed here?

  • @Albarjakmajd
    @Albarjakmajd 2 роки тому

    amazing!

  • @RiccardoVecchi1
    @RiccardoVecchi1 3 роки тому +1

    Sorry can you please give me the R code for the crude rate? I can't find it

  • @anmolpardeshi3138
    @anmolpardeshi3138 3 роки тому

    if Poisson regression (at basics) is a method to model counts then how is person-time coming into play here. For example: what if I want to run Poisson regression for a die throw. The counts on the upper face of the die could also be modelled this way right? In die throw there is no person-time involved. So, my basic question is - how is Poisson regression transitioning from simple models to these person-time involving models. Pls let me know. Thanks

  • @newname3201
    @newname3201 3 роки тому

    This is very good!

    • @newname3201
      @newname3201 3 роки тому

      simple to follow and didn't waste any of my time! I just want to take this chance to say thank you and have a wonderful day! You are amazing

  • @ranjanibhat4417
    @ranjanibhat4417 4 роки тому

    Can i use poisson regression for the data containing cases and deaths due to a disease year wise

  • @michaelvendredi8274
    @michaelvendredi8274 4 роки тому

    Thank you!

  • @soumitrakar3998
    @soumitrakar3998 Рік тому

    Can you kindly add the Overdispersion and excessive zeroes and two Part videos of Poisson regression of Mary Clare Kennedy also a part of the video lecture series of Poisson regression then it will be helpful for viewers.

  • @raitup00
    @raitup00 4 роки тому

    Where we can get the database?

  • @sebastianbalbo1906
    @sebastianbalbo1906 3 роки тому

    I doubt that those kind of STATISTICS MODELS COULD BE USEFUL IN PARTICULAR WHEN YOU HAVE TO UNDERSTAND THE EFFICACY OF PHARMAS

    • @billbentley3
      @billbentley3 2 роки тому

      these are exactly the kinds of models used for that.