Control Chart for Multiple Products – Normalize?

Six Sigma – iSixSigma Forums General Forums Tools & Templates Control Chart for Multiple Products – Normalize?

Viewing 3 posts - 1 through 3 (of 3 total)
  • Author
  • #211626


    Hi everyone,

    I’m sure I’m not the only one, but I work in an industry where one line produces multiple products with multiple targets for the same variable. So for example, a box factory that makes different types of boxes: a spec for “length of a given side” exists for every product, but each type of box has a different spec for that.

    I’m wondering if it would be worthwhile to normalize the data to the target for that particular product so that the overall process for meeting that spec can be evaluated.

    I want to study the % deviation from target of the overall process. Part of the reason for this is that I have limited access to lab data (very lean lab).


    Mike Carnell

    ChemEDad If all you want to know is if the products run differently you can do that with a hypothesis test. Test for Normality (do not transform the data if it is not normal it doesn’t have to be), Test for Homogeneity of Variance (Equal variances) and then One Way ANOVA or t- test. If you do this correctly and calculate your sample size and select the amount of change you need to see to be critical you will get a definitive answer with a known alpha and beta risk on your decision. Do the chart and it is based on your interpretation.

    It has been a while since I worked on a box making machine but when you change size you change setup. That by itself should change things but you never know until you check the numbers.


    Chris Seider

    @mike-carnell gave you good advice.

Viewing 3 posts - 1 through 3 (of 3 total)

You must be logged in to reply to this topic.