DOE Excel

Пікірлер: 16

  • @gonzaloallendes1078
    @gonzaloallendes10783 ай бұрын

    Great, thanks. You can also use a neat AVERAGEIFS function in your mini matrix at 17:20

  • @gerryberard3231
    @gerryberard32314 жыл бұрын

    Very nice example. Thanks for sharing.

  • @dotbstudios
    @dotbstudios7 жыл бұрын

    Well done!

  • @firstnameylastnameo4848
    @firstnameylastnameo48484 жыл бұрын

    Thanks for your video. I would like to design an experiment with 3 factors but each factors have many levels, not necessarily with the same number. How can I do that?

  • @ramdasmenon0908

    @ramdasmenon0908

    2 жыл бұрын

    The simplest way it to use the generalized factorial option in Minitab. The R Programming Language also has several DOE packages that you can use. You can do this in Excel also, though it will take some doing.

  • @paawantoor1
    @paawantoor18 жыл бұрын

    why the factors ABCD has 1 and -1 factor, usually it is 0,1 and -1..?

  • @adrianvega9946

    @adrianvega9946

    4 жыл бұрын

    Because not using centerpoints. Using a centerpoint is to find out if there is a curvature. And if we intend to use linear model then isn't needed.

  • @paawantoor1
    @paawantoor18 жыл бұрын

    to find out quantile, why it was first 1/20 then 1/10+(1/20) then in residuals it was 1/32 and 1/16+(1/32)? reason for selecting factor 20 and 32?

  • @Underthelab

    @Underthelab

    6 жыл бұрын

    hey there, thank you for your useful video, kindly could you assist in the question by Rose. Your assistance is appreciated

  • @vincewest5790

    @vincewest5790

    6 жыл бұрын

    In looking at the effects, there were 10 effects, four first order effects and 6 second order effects. Normally distributed these should fall into 10 equally divided quantiles, hence the 1/10 increment. The first quantile being 0-10, the mid point is 5 percent or 1/20, and the next is 15 percent, or 1/10 + 1/20, and so on. For the residuals, there were 16 points that if normally distributed should be in 16 equal quantiles. Using similar logic the first midpoint is at 1/32, and the second quantile midpoint is at 1/16 + 1/32. Hence these numbers are just based on the number of normally distributed points to evaluate.

  • @svaranasi542

    @svaranasi542

    5 жыл бұрын

    @@vincewest5790 This is very great inf, kindly please tell abt the calc

  • @chandrashekharb8073

    @chandrashekharb8073

    5 жыл бұрын

    @@vincewest5790 But she sorted the effect list from smallest to largest, why didnt she do the same for the quantiles?

  • @vincewest5790

    @vincewest5790

    5 жыл бұрын

    Chandrashekhar B, I don’t remember the video exactly but you have to sort the effects so you can put the into quantiles which are ordered from smallest to largest by definition

  • @Monkeysal07
    @Monkeysal07 Жыл бұрын

    why are the beta coefficients divided by 2 ?

  • @jessetello7620

    @jessetello7620

    22 күн бұрын

    im in her class. she never explained it.

Келесі