Statistical Process Control
3 sigma, 6 sigma
.26% defective, 3.4 defects for every million produced
Variable Data
Can be measured using a continuous scale, x bar and R charts
For Cp and Cpk if more than 1 then
Capable
Type II Error
Continuing to think everything is fine when a change has occurred (ex. mean may have shifted)
P charts
Designed to track proportion defective in a large sample
For Cp and Cpk if less than 1 then
Incapable
Characteristics of design specifications
Preset by design engineers to define acceptable ranges Based on customer expectations and how the product works (not statistics!)
Statistical Process Control
SPC, used to determine whether a process is performing as expected, goals are predictability and repeatability, SPC is an auditing procedure.
Attribute level data
Simply cataloged by descriptive and discrete characteristics, p charts and c charts
C Charts
Used to track average number of defects per single unit of output
Centered Process (Cp)
Where the mean of the process and the middle of the specification limits are equal
Type I Error
Willingness to think something's wrong when it's actually not (False Alarm)
Control Charts for Variable Data
Mean (x-bar) charts track the average performance (central tendency) over time. Range (R) charts tract the variation of performance (distribution) over time.
Un-centered Process (Cpk)
Mean of the process and the middle of the specification limits are not equal
Process capability
Measure of the ability of a process to meet preset design specifications and determines whether the process can do what we are asking it to do.
Barely Capable
Specification limits and process variability are equal; Capability Index --Cp equal to 1.
Incapable
Specification limits are narrower than process variability; capability index--Cp less than 1.
Highly Capable
Specification limits are wider than process variablity; Capability Index-- Cp greater than 1