Z Bench Calculation

Six Sigma – iSixSigma Forums Operations Call Centers Z Bench Calculation

Viewing 3 posts - 1 through 3 (of 3 total)
  • Author
  • #54702


    I’m new to this forum. Can somebody explain the difference among Zbench, Zst and Zlt. I’m working on a project where targeting to improve the number of sales hit by agents. To start with I have mean = 21 sales/month/agent, Std dev = 11 sales/month/agent, Specification USL >= 3. Can someone help me with Zbench calculation?


    Chris Seider

    Basically look at a normal distribution table in a stats book. The percent defective aligns with your Z. If you use excel to clarify, Z = -1*normsinv(defective) (defective) is a decimal with 0.5 being 50% defective.

    Zst and Zlt and convoluted topics for a forum like this but consider Zst the best you could ever achieve with no special causes one could expect. Zlt is the sigma level for your process with the special causes included.

    Most industries one can get a Zlt with 3 months of data but some industries such as chemical and food often require a year’s worth of data (because of known seasonal effects) to get best estimate for Zlt but most just use 3 months or less for ease of data gathering.

    Sigma level is defined as Zst.

    I would worry on reducing the ppm or % defective and not worry as much about which Z you have. If you want to see much past discussion on how much difference between Zst and Zlt–I still remember the best advice I learned long ago–“since we know all processes are different, MEASURE the difference.” But many folks don’t want to bother with all the work so they often use some standard number across an organization.


    Chris Seider

    I tried to edit last response but to no avail.

    Look up past discussions on Zst vs Zlt on this site.

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

You must be logged in to reply to this topic.