• StuW replied to the topic Normality in the forum General 12 years, 8 months ago

    Here are some ideas:
    * Plot all the data, and look at the histogram.  Check the shape.
    * Plot all the data using a group color, or symbol, and check the overall group shapes with the various groups identified.
    * Run a boxplot for each group to look at the mean, quartile points, and any outliers for each group.
    * Run the 1-Way ANOVA, and ch…[Read more]

  • StuW replied to the topic 2 sample t test in the forum General 12 years, 8 months ago

    Nobody has asked ajl if these experimental units can be related in some way, such that the actual approach should be a paired t-test?   In that case, equal sample sizes becomes the requirement, and the comparison is made on the difference of measurements made on “like” units.   Given the lack of statistical knowledge that is exhibited in m…[Read more]

  • There are some major factors to consider with this.   First off, is the upper management support for the training in place?  Without it, you won’t have the commitment in people, or time, to do this effectively.  Secondly, this has to be aligned with the key project objectives of the organization, so after training the personnel can begin the pro…[Read more]

  • StuW replied to the topic stats in the forum General 12 years, 8 months ago

    It is also difficult to know exactly what has been taught to students when these questions arise.  In this case, do they know the Normal approximation to the binomial distribution for larger sample sizes?   In that instance, the solution becomes rather quick by stadardizing the 56% value (or 44%, value if considering wins) and then finding the ar…[Read more]

  • StuW replied to the topic Control charts in the forum General 12 years, 8 months ago

    Don’t confuse control and capability concepts as they are distinct concepts!  
    Control charts assess stability and whether, in this case, your 25 weekly CSAT readings represent a consistent process, with a mean of about 0.446.  Check the chart.  Does it show trending, either up or down, or any other evidence of non-stable behavior?  Are…[Read more]

  • StuW replied to the topic Gauge RR in the forum General 12 years, 8 months ago

    There are several reasons that gauge R&R should be less than 10% for most gauge applications.  In some texts, I have seen P/T ratios between 10 and 30% classified as “marginal” gauges.
    First, the disposition of units is subject to incorrect classification as the ratio rises.  This can lead to issues with sending on product that does not meet your…[Read more]

  • StuW replied to the topic DOE-attribute in the forum General 12 years, 9 months ago

    Compute the percentage pass (or fail) for each cell, combination of -1 and +1 factors, in your design.   Use that result as the dependent variable in the analysis.   The factor level combinations are your set of independent variables in the matrix. 
    Several points, some of which were covered with the previous responses.   If you have too few uni…[Read more]

  • StuW replied to the topic Two-Way ANOVA vs GLM in the forum General 12 years, 9 months ago

    As pointed out, in the balanced case these analyses are exactly the same and should provide the same results.   What does change is the output provided from the software which is often different depending upon whether the 2-way ANOVA or GLM routines are run.   In the balanced case running the analysis in both ways may give you a par…[Read more]

  • The more the inherent risks are to YOUR business, the more it is incumbent upon your organization to take steps to avoid these defects.  As an example, in the semiconductor business many issues attributable to our suppliers such as photoresists, silicon substrates, or gasses used will be blamed on our company.  In fact, we had a supplier problem w…[Read more]

  • The classic text “Statistical Process Control” 4th edition 1972, by Eugene Grant and Donald Leavenworth has a table in the Appendix, table B, which has d2 and c2 values in increments of 5 between 25 and 100.   The c2 values convert from sigma-bar to an estimate of sigma using “n” in the denominator, not “n-1”.  You can easily interpolate to get…[Read more]

  • StuW replied to the topic before and after data. in the forum General 12 years, 9 months ago

    Your main issue here is that the pre to post test comparison should show improvement since you are providing instruction to people that are taking the class presumably because they don’t have those skills, correct?  So then the question becomes how effective was your training?   In other words, your scores should improve by at least tw…[Read more]

  • StuW replied to the topic Metrics Creation in the forum General 12 years, 9 months ago

    Sounds like you would need to create an employee morale survey or use one already established to assess this.   There are companies that specialize in these types of Surveys, but if you are well versed in survey design creation, you could also create your own.   Many of these start with a baseline to assess where employee satisfaction is now,…[Read more]

  • It is somewhat difficult from your description to tell exactly what test is being conducted.  The calculator provided in the reply given is useful if you are measuring a percentage, but I’m not sure that is what you are measuring.   Some questions that you need to answer to process this correctly:
    – Are the units “paired” or independent sa…[Read more]

  • Unfortunately, a two sample test of proportions from Minitab’s calculator shows sample sizes of n=376 per group, and that is to see a difference from a 4% null hypothesis to an alternative rate of 1% on the new design (alpha=0.05, power=0.75).   For 2%, the sample size jumps to about a 1000 per group so unless you are fairly certain the new de…[Read more]

  • StuW replied to the topic To Z or not to Z? in the forum General 12 years, 9 months ago

    Darth:  That is a good exercise because Cp and Cpk are both driven by the sample variability in the estimate of sigma.   My recollection is a spread of +/-0.2 on Cpk for sample sizes of about 30.  This is often overlooked when a Cpk is determined and then immediately converted to a defect level.
    MBBinWi:  I was responding to provide some addit…[Read more]

  • StuW replied to the topic To Z or not to Z? in the forum General 12 years, 9 months ago

    A couple of points:
    Cpk does include both ends of the distribution, by definition it is min(Cpl,Cpu).  Your complaint is more that it does not capture probabilities from both ends, which is not the purpose of Cpk to begin with.  Cpk is a figure of merit on whether a distribution is capable to the spec limits.   It started with 1.0 as…[Read more]

  • First off, what is the approximate leak rate that you are seeing?   That drives required sample sizes in the go-no-go realm of measurement.  The higher the rate of failure, the easier to remain in the go-no-go measurements, while as the rate drops down it would be much more convenient to have a continuous measure to rely upon. 
    Since you sugg…[Read more]

  • StuW replied to the topic To Z or not to Z? in the forum General 12 years, 9 months ago

    There are some subtle advantages to keeping the concepts of standardization and capability indices somewhat different.  From a teaching perspective, one can introduce the concept of standardizing a value from a N(mu,sigma^2) distributionto a N(0,1) distribution in order to determine how many standard deviations a value is from the mean, without e…[Read more]

  • SAM:
    Here is the general guideline for a paired t-test versus a 2-sample test.  If the measurments are made on the same experimental units, and those units are expected to differ from one another in a material way, then the paired t-test is appropriate, as it removes a source of variation (that due to the units themselves) and allows the test to…[Read more]

  • Tom,
    Why would you believe the process you are measuring should be restricted to have an underlying population average that corresponds to your gage discrimination?   Of course you can report an average that is whatever the data suggests it is.
    The issue you may have is a lack of resolution in your measurements for detecting changes in the p…[Read more]

  • Load More