2: Dealing with missing data

Поделиться
HTML-код
  • Опубликовано: 21 окт 2024

Комментарии • 12

  • @ЮлияМузыченко-х5п

    Hello, thank you for the video! Do we do analysis of the patterns for any mechanism of missingness? If yes, what's the point of analysing type of missingness if we will run imputation?

  • @Stelnice
    @Stelnice 3 года назад +1

    Hi! Our statistician told us that we can only use listwise, pairwise, imputation with experimental research. I'm just a little confused because the example is about the predictors of homeownership (age, education, etc.). Does that mean we can also use it for social research etc? Thank you!

  • @janinajochim1843
    @janinajochim1843 4 года назад +2

    Hi :)
    Very well explained video, thank you! I am working with a dataset that has over 1.500 variables, however I am only using about 25 variables for my analyses.
    Is it ok if I run Little's test only over the variables that I end up using for my analyses? For one variable, I am getting an error that says "sample covariance matrix is singular for at least one missing-value pattern" -- do you have any recommendation as to how to handle that? Thank you in advance!

    • @mrr1491
      @mrr1491 2 года назад +2

      Did you ever solve for your error? I"m encountering the exact same error right now and I'm struggling to find responses for how to deal with this.

  • @Blupblup100
    @Blupblup100 4 года назад +1

    Is is possible to impute data that was collected retrospectively?

  • @RamonGeorgeAtentoPhD
    @RamonGeorgeAtentoPhD 4 года назад +1

    Hello from the Philippines. I have a question here

  • @nishantthakur691
    @nishantthakur691 3 года назад

    Actually, I am having some difficulty to analyzed the hazard ratio from digitized Kaplan Meier data. I shall be thankful to you if you please help me.
    Thanks in advance

  • @jamespaz4333
    @jamespaz4333 4 года назад +2

    Most tutorials recommend imputing missing values with mean or median 🤔

  • @Silverwing_99
    @Silverwing_99 3 года назад

    You lost my respect after you said you use spss 😿 - nice explanation though

    • @pabgrg3199
      @pabgrg3199 3 года назад +2

      the important thing is you should have a very good understanding of the test you are using, whether u use R/STATA/SPSS/MATLAB/SAS.....I have seen so many self proclaimed R users who don't even understand the basics of stats