Downloaded from https://www.everyspec.com MIMiDBK-683(AT) TABLE II-4. to~sx~ G Pitfall Not verifying calibration of —-=* =W@-nt. JMZ41 a-Wable parta as defective when they are good or good uhen they are defective. Failing to verify testing ~*s by Inspection Capability Studies. a. Judging accep~ble parts as defective or defective parta as acceptable. b. Conflicts over what Ocmutitutea good and defective -. c. QMlflicting deciaiom an to whether proceoa ia in contxml. Not selecting the proper Control Chart for the characteristic being measumd. Misleading results; looking for assignable causea when there are none and failing to look for assignable cause when one is present.
UJing subgroupe sizes of 2 or 3 (Xaxni RCharta) for an ine~neive inspection. Increased chance of not detecting a p~ee shift when 0~ b occurred. U8ing @group sizes of 10 or more (X and R Charts). ~ncrea8ed inspection costs becauee X and R Charts are not ae efficient for large group size (n). Using small subgroup sizes for a p chart with low percent defective. Control limits are such that the chart cannot detect an out of control condition. Qmfuaing subgroup sample size with.the number of subgroupo when selecting Az, k, D~, & factors for control li.mits- Incorrect control limits resulting in failure to detect assignable cauees or looking for assignable causea when none pmment.
Not selecting 8UCC088iV9 unite for eubgroups (~and RChartS). Assignable causes ~ part of within subgroup varid ion cauaing unneceaearily wide control limits. This nesults in failure to detect assignable causes when they are present. 43