SPSS (R Essentials) - Zero-Inflation Count Models

Поділитися
Вставка
  • Опубліковано 12 січ 2025

КОМЕНТАРІ •

  • @tegaxel
    @tegaxel Рік тому

    Thank you very much Dr. Perret for your helpful video (I have been looking for valuable guidance like yours for months). If it is possible I would like to ask you just one question so that I can apply the procedure correctly: if I use only one independent categorical variable (with 3 levels) and one dependent variable, keeping the tick in 'use count predictors for zero model', what is the difference between the output table 'count model coefficients' and 'zero-inflation model coefficients' (i.e., which one should I consider and how should I interpret them)?
    Thanks for your help and great expertise

  • @mackmariani
    @mackmariani 4 роки тому +1

    This is very helpful, but I do have one question: how do you generate an R square for the models?

    • @mackmariani
      @mackmariani 4 роки тому +1

      From what i've been told, I need a deviance statistic, not an R square for poisson

  • @รชนีกรแซงโคตร

    Hello, I'm from Thailand. I would like to ask you what version of this program?

    • @jensk.perret6794
      @jensk.perret6794  2 роки тому +1

      The video used SPSS 23 but it looks almost identical in more current versions.

    • @RqueErre
      @RqueErre Рік тому

      @@jensk.perret6794 Same problem, I'm using the last version and there's no such option for zero inflation data

    • @birtukangeleta3978
      @birtukangeleta3978 Рік тому

      Hello I'm from Ethiopia. is it possible to use run double hurdle model on SPSS? Eventually Thank You!

  • @Cyrribrae
    @Cyrribrae Рік тому

    I keep referring back to this video and the impression I keep getting is that this just lacks a lot of functionality haha. Not having a real test of model or goodness of fit (right?). I saw in other places you get confidence intervals from bootstrapping.. but that doesn't seem to be an option here? But guess this probably doesn't get enough use to warrant more development from IBM?

    • @jensk.perret6794
      @jensk.perret6794  Рік тому +1

      I would fully agree, but this function not yet being an SPSS core function might mean that it will take a long time for it to be fully usable. I would also recommend to get standard error via one way or another and then calculate a confidence interval.

  • @PeeKayTamu
    @PeeKayTamu 4 роки тому +1

    hello sir, can you please help me. How do i install R PSCL package when i use R essential. I am using SPSS 23 and following your video how to install R essential with R 3.1.0. However, I could not use the zero inflation count model. It says R PSCL package is required. Please HELP, I would be highly grateful. I am a master degree student and learning it for my educational purpose. Thank You.

    • @jensk.perret6794
      @jensk.perret6794  4 роки тому +1

      The easiest way would be via the R menu. At 'Packages' select 'Install Packages', select a mirror and then look up the corresponding package. Alternatively look up the package at CRAN homepage and install it manually.

    • @PeeKayTamu
      @PeeKayTamu 4 роки тому

      @@jensk.perret6794 thank you