Added-variable plots/partial regression plots

Поделиться
HTML-код
  • Опубликовано: 9 янв 2025

Комментарии • 17

  • @Manny123-y3j
    @Manny123-y3j Год назад +1

    This was a really clear explanation and good example. I understand it now. Thank you!

  • @sethozment2903
    @sethozment2903 3 года назад +3

    Better than any textbook explanation I've seen, thank you so much!

  • @danxx0090
    @danxx0090 3 года назад +1

    Thank you so much for this video! It has solved my very important doubts that I couldn't find anywhere else.

  • @ΣτέλιοςΚαραλής-ο5λ
    @ΣτέλιοςΚαραλής-ο5λ 2 года назад +1

    Most helpful explanation. Thank you very much!

  • @aileen55
    @aileen55 2 года назад +1

    Thank you for the video!!! Currently studying stats at ucla. This helps me round out the math with a conceptual, intuitive understanding

  • @nasheedjafri3564
    @nasheedjafri3564 3 месяца назад +1

    What if you get a steep negative slope line in your added variable plot?

    • @statisticsninja
      @statisticsninja  3 месяца назад

      @@nasheedjafri3564 If a variable has a steep slope positive or negative then you want to include it in your model.

  • @kuku7330
    @kuku7330 2 года назад +1

    so helpful thanks so much

  • @larissacury7714
    @larissacury7714 2 года назад +1

    Hi, thank you! Is it equivalent of doing a partial correlation plot? I'm having some trouble finding an easy way to plot this partial correlation: pcor.test(data$X1,data$X2,
    data$Z method = "spearman") .... Z is the controlled variable, which influences both X1 and X2 (I'm using R studio) Ps: why are all variables centered?

    • @statisticsninja
      @statisticsninja  2 года назад

      These are similar, but different. I would use car::avPlots() if you need scatterplots. Would a bar plot of the estimates from ppcor::pcor() work? Centering and scaling the data helps make it easier to evaluate the magnitude of the values.

  • @WhySoBroke
    @WhySoBroke 2 года назад +1

    This was great! Where can we get the doc or written instructions?

    • @statisticsninja
      @statisticsninja  2 года назад

      I put the files in the shared files section of my website. www.statistics.ninja

  • @TheBugfoot
    @TheBugfoot 4 года назад +2

    thanks

  • @raoulroberts2024
    @raoulroberts2024 4 года назад +1

    Are you using R or Stata?

  • @samuelboff3256
    @samuelboff3256 3 года назад +1

    I understood it well until you use American football as an example.

    • @statisticsninja
      @statisticsninja  3 года назад

      Okay how about this analogy. If I have an all-star goalie, would it improve my team to get another all-star goalie? Predictor variables with flat added-variable plots have little to add to the model. Adding such a variable to a model does not improve the model. Much like double filling a position does not improve a team.