Previous   |    Next

doi:10.2204/iodp.proc.308.205.2008

Reproducibility

We used samples from whole-core trimmings to run repeat experiments from whole-core trimmings at Penn State to gauge the reproducibility of our experiments. We conducted two types of tests: multiple hydrometer runs on a single sample and hydrometer runs on a single sample in which we varied the initial dry mass. We show an example of each type of test in Figures F10 and F11.

Samples were prepared by “cone and quartering” to ensure homogeneity. In this process, sediment is mixed and piled into a cone that is then divided into quarters, and opposite quarters are combined.

The first-order results of our experiments show that particle size distribution curves from both types of tests generally match each other (Figs. F10, F11). A statistical treatment of the reproducibility experiments was conducted by the Statistical Consulting Center at Penn State. They applied a cubic-spline regression method and concluded that our experiments are fairly consistent and reproducible.

We did not test the reproducibility of the experiments performed at MIT. However, the equipment, materials, and procedures used at MIT were very similar to those used at Penn State. Furthermore, we distinguish the MIT experiments as red squares in Figures F6 and F7 to show that the MIT experiments lie within the expected range of results.

Top of page   |    Previous   |    Next