Font Size
Topic Question About Significance

Question About Significance

Home Forums General Forums Tools & Templates Question About Significance


This topic contains 7 replies, has 5 voices, and was last updated by  S Nandy 1 year, 1 month ago.

Viewing 8 posts - 1 through 8 (of 8 total)
  • Author
  • #592215 Reply

    If I’ve % (Y) and my X’s is also in % .For Hypothesis what test should I run in Minitab to check the significance ? Do I need to convert % data in some other format ?

    #592220 Reply
    #592502 Reply

    Thanks Chris !! Correct me if I’m working I’m starter . % is attribute data , Right ?
    Do I need to convert % data in some other format to run hypothesis test ?

    #592585 Reply

    Hi @snandy,

    Percentage is a form of attribute data, yes. You do not necessarily need to convert it into another format to run a hypothesis test. The famous answer is; it depends. It depends on what you’re testing and what you’re looking to find from it.

    During my black belt course with Process Management International (UK based), they handed this table out which is very helpful on hypothesis test selection. See attachment and I hope it helps.

    #592587 Reply

    This time with attachment, hopefully?

    #592951 Reply

    @jamesm – it’s a good flow chart but what you need to remember is that the “rules” in that flow chart are, of necessity, ultra conservative and every one of them has multiple caveats.

    For instance – the split on the decision concerning normal vs non-normal data. t-tests and ANOVA are quite robust with respect to non-normality and will usually give you the same results as the Wilcoxon-Mann-Whitney and Kruskal-Wallis tests.

    The same is true for ordinal – whether you use non-parametric or parametric methods for analysis of this kind of data will depend on the question you are trying to answer.

    This is not to say that the tree you provided is useless rather it is to warn you that the tree is set up to try to insure that someone unfamiliar with the world of statistical analysis will not make serious mistakes when they are trying to analyze data without any additional outside resources such as a statistician. In short use the rules but never lose sight of the fact that they present limitations with respect to analysis that, in general, don’t exist.

    #593174 Reply

    % is not necessarily attribute. I’ve seen some “teach/mentor” that and it’s just wrong. % completion, % correct, etc (transactional minded) can be handled as continuous data (especially if plenty of precision).

    Conclusions on continuous data is always “easier” but that’s not why I’m saying % can often be used as continuous data.

    #602474 Reply

    Thanks for the discussion . I got it now .

Viewing 8 posts - 1 through 8 (of 8 total)

Register Now

  • Stop this in-your-face notice
  • Reserve your username
  • Follow people you like, learn from
  • Extend your profile
  • Gain reputation for your contributions
  • No annoying captchas across site
And much more! C'mon, register now.

Reply To: Question About Significance
Your information:

5S and Lean eBooks

Six Sigma Online Certification: White, Yellow, Green and Black Belt

Six Sigma Statistical and Graphical Analysis with SigmaXL
Six Sigma Online Certification: White, Yellow, Green and Black Belt
Lean and Six Sigma Project Examples
GAGEpack for Quality Assurance
Find the Perfect Six Sigma Job

Login Form