Dummy Coding JASP

Поділитися
Вставка
  • Опубліковано 3 гру 2024

КОМЕНТАРІ • 5

  • @marie-lauredecaster5667
    @marie-lauredecaster5667 3 роки тому +2

    Hey! How do you put a nominal variable in the covariates? (min 05:44) In my analysis it doesn't work because I have to import a cont one? (see the ruler in de right bottom cornor) Kind regards!

    • @danielshowalter-mathematic4316
      @danielshowalter-mathematic4316  3 роки тому

      Can you explain a bit more? These are nominal variables that I'm using as covariates (but I can only do so because the nominal variables have only two categories -- if they have more, I need to dummy code first). For continuous variables (ruler icon), you should be able to put them directly in without worrying about dummy coding. Ordinal variables are the tricky and controversial ones -- technically, the best thing to do with those is to run a separate type of regression for ordinal variables.

  • @ICQwhereRU
    @ICQwhereRU 3 роки тому

    interpretation is not correct, you need to compare to the reference category (the omitted one for the gender)

    • @llewmills6406
      @llewmills6406 2 роки тому

      Yes I wondered about this as well. I would have thought the coefficient for intersex was intersex vs male and the coefficient for female was female vs male.

    • @llewmills6406
      @llewmills6406 2 роки тому

      Unless JASP does type-3 sums of squares coding by default.