Footnotes on the Measurements
(1) Supplier: Specification of a clean new roll given by the supplier of the anilox roll.
(2) Capatch: Each result represents 1 measurement, in total 4 measurements per anilox roll. Capatch low volume (green) strip used on anilox roll number 1 and 2. Capatch medium volume strip (blue) used on anilox roll number 3, 4 and 5. So 1 result is only 1 Capatch measurement.
(3) TCS: Confocal Laser Scanning Microscope. Each result represents the average of 5 measurements, in total 20 measurements per anilox roll. So 1 result is the average of 5 TCS measurements. *
(4) Volucheck: Each result represents the average of 3 measurements, in total 12 measurements per anilox roll. So 1 result is the average of 3 Volucheck measurements. **
(5) Urmi II Each result represents the average of 4 measurements, in total 16 measurements per anilox roll. So 1 result is the average of 4 Urmi II measurements. **
(6) Average per anilox roll of 4 results of Capatch (total of 4 measurements), TCS (total of 20 measurements), VoluCheck (total of 12 measurements) or URMI II (total of 16 measurements).
* This means that the variation of these results is much less than the variation of the Capatch results because of the way in which these results were obtained, and not because of lower reliability of the Capatch.
** This means that the variation of these results is much less than the variation of the Capatch.

Validity and Reliability
The variance of means (TCS, VoluCheck and Urmi II) is always smaller than the variance of unit measurements (Capatch). The data as given appears to be biased against the Capatch. We would like to have an additional comparison test done by an independent institute so that unit or equal sample size means are compared between devices in order to make a valid comparison of the reliability of the respective measurement techniques. This comparison shall be complicated by the fact that the various techniques are taken across differing surface areas, so that inherent differences in variance shall occur. On the other hand, each act of taking a measurement is the relevant comparative unit for the user.
