I85PREP
⌘ K
CFSQA study materials
CFSQA study resources

CFSQA: Quality Systems, Measurement, and Continuous Improvement

CFSQA · Work through your exam, one topic at a time.

All guide sections

Quality Systems, Measurement, and Continuous Improvement

This domain tests whether you can audit a quality system as a system: document control, calibration, complaints, nonconforming product, corrective and preventive action, measurement system quality, and the statistics used to judge process capability. Expect questions on CAPA effectiveness and control-chart interpretation.

System elements auditors verify

Documented procedures and records under revision control; training linked to competence; calibration traceable to standards; nonconforming product and rework controls; complaint trends; supplier approval and performance; management review with inputs and outputs; internal audit programme with follow-up.

Measurement

  • MSA / Gage R&R: %GRR under 10% generally acceptable, 10–30% conditional depending on application, over 30% unacceptable for the measurement's purpose.
  • Control charts: variables (Xbar-R) and attributes (p, np, c, u). Control limits come from the process, specification limits from the customer - they are not interchangeable. Points beyond limits or non-random patterns indicate special cause; over-adjusting common-cause variation makes it worse.
  • Capability: Cp = (USL − LSL)/6s; Cpk = min(USL − x̄, x̄ − LSL)/3s. Cpk adds centring.

Root cause and CAPA

5-Why, fishbone, and FMEA support root cause. Corrective action removes causes; correction fixes the instance. Effectiveness verification is a separate, later step. "Retrain the operator" is rarely a root cause.

Worked example

Fill volume limits are 8–12 mL, mean 10.6 mL, s = 0.5 mL. Cp = 4/(6×0.5) = 1.33; Cpk = min(12−10.6, 10.6−8)/(3×0.5) = 1.4/1.5 = 0.93. Reasoning: the spread is acceptable but the mean is off-centre, so capability is poor (0.93) and the fix is centring, not reducing variation. If the same gauge gave %GRR of 28%, measurement error is conditional and may not support release decisions at that tolerance.

Traps that cost marks

  • Closing a CAPA when the action is implemented rather than when effectiveness is verified.
  • Mixing Cp and Cpk, or control limits with specification limits.
  • Accepting "operator error" as a root cause.
  • Auditing calibration certificates for presence without checking range and traceability.
  • Ignoring complaint and environmental monitoring trends that show drift.