About Todd Dauterman Expertise I can assist you with: Facility Layout (including machinery) and Design; Construction; Process Automation, Production and Inventory Control; Machinery Engineering (design and modifications also); Logistics (shipping and receiving); Facility Management; OSHA/EPA Programs; Payroll; P&L 50 Million
Experience Business owner for the past 9 years. Plant Manager ? 8 Years; Senior Plant Engineer ? 2 Years; Production Manager ? 2 Years; Industrial Engineer ? 2 Years
Expert: Todd Dauterman Date: 9/16/2003 Subject: inference population from sample
Question we have a plant.We buy some parts of our production and we don't know anythinng about the proportion of it's defective.These parts are in a huge number(about milion).Now I want to know how can I take a small sample from it and then I inference it's result for population?and how can I calculate the probability of it's error?what relation is between sample size and the probility of get an error inference?
Regard.
Azadeh baniasad.
Answer I understand your scenario but you are asking yourself the wrong question and here is why. When you buy a carton of eggs from the store, you expect all of them to be there in usable form. It is the same for you buying sub units from your vendor. You need to read up on QS-9000. QS 9000 is were you dictate the vendors quality. It worked very well for the US automotive manufactures (assembliers). If the sub unit is complex the I would suggest that you find a vendor that utilizies Six Sigma. If they are performing to the Six Sigma standards correctly you will end up with one defective part per million units.
