The variability of the protection factors for sunscreening products

The variation of the preotection factors for a sunscreening product as determined for different test subjects can be described by a logarithmic normal distribution. It is therefore more appropriate to use logarithmically transformed data for statistical analyses. By means of an analysis of variance it was shown that the influence of the subjects is responsible for only 12% of the total variance, whereas the remaining 88% is due to experimental error. Therefore, the difficulties in determining precise protection factors must be ascribed to deficiencies in the method itself. It was further concluded that correcting the protection factors with respect to a simultaneously measured reference preparation is not a means of improving the precision of the data. The reference substance, if it is included in every measuring series may, however, be used to check the long-term constancy of the method and to compare results from different laboratories.