Letter to the quality engineering editor

Comment on the Gauge R&R Literature Review by Soares et al. (2022) I read with interest the above paper that was a systematic literature review whose stated aim was “to assess the state of the art in Gauge R&R [measurement assessment] studies.” Given that goal, I was disappointed that the review paper failed to even mention any of the following proposed improvements to the traditional design and analysis of gauge R&R measurement assessment studies. This includes, so called, augmented assessment plans, utilizing baseline data, and selecting parts for the measurement study using leveraging from the baseline (Browne 2009a,b; Browne et al. 2009a,b, 2010; Stevens et al. 2010, 2013, 2015). These ideas provide much more efficient measurement assessment studies with no increase in cost. As such, they should be considered as desirable improvements to the traditional Gauge R&R studies. To make a simple comparison, consider the traditional Gauge R&R study where 10 parts are selected at random from the production process and measured 6 times each (here we assume an automated measurement system, so we don’t consider operators). If we also assume we have a large baseline of once measured parts (commonly freely available since these data are often collected for some other reason), an alternate measurement assessment plan is to use leveraging and select 10 parts with either large or small values from the baseline and measure them each 6 times. Comparing these two plans (that require the same number of total measurements in the assessment study) we see that the leveraged plan (that takes into account the baseline values for the selected parts) is much more efficient than the traditional gauge R&R plan. The reduction in standard deviation of an estimator for the ratio of the measurement variation to the overall variation is between about 10–45%, with larger reductions when the actual measurement variation is larger.