The plot of all data using the LAS-W technique is shown in Appendix E. Clearly, there is significant, although not large, scatter in this data. As shown in Table 6 (columns labeled "all data"), this scatter is reflected in the calculated median (50 % of mass less than) diameters. As discussed on Section 3.1.2, the sets B, J, N, and U were found to be outliers and it was shown that the correction factors are not enough to correct them. Therefore, the median was recalculated without these sets. Table 6 shows how the median diameters and their standard deviations change with the elimination of the outliers. The standard deviation is reduced but not for all cements. For instance CCRL 131 median size did not change. By looking at the graphs in Appendix E, it is clear that different outliers could be picked if instead of cement 114p another of the other cements was taken as the reference. As no details are known from these measurements, it is hard at this time to determine why.
| Table 6: Average median diameter from LAS-W method. | ||||
|---|---|---|---|---|
| Cement | All data | no B J N U | ||
| Median Diameter [µm] |
Standard Deviation [µm] |
Median Diameter [µm] |
Standard Deviation [µm] | |
| 114P | 12.9 | 2.6 | 13.3 | 1.8 |
| 131 | 10.3 | 2.4 | 10.2 | 2.4 |
| 132 | 8.0 | 2.4 | 7.3 | 1.1 |
| 135 | 12.2 | 1.7 | 12.1 | 1.4 |
| 136 | 12.9 | 2.0 | 12.7 | 1.6 |