Next: Conclusions Up: Prucha, etal.: Reproducibility Previous: Testing reproducibility

Results from the last report

The testing of SEP-100 was completed before it was made available as an electronic document last spring. All testing was carried out on an IRIX 6.5 operating system. The ER testing was done first. The results are shown in the following table.

Testing Round No. of ER figs No. that failed % that passed

1 95 7 92.6%

2 93 2 97.8%

3 93 100.0%

Testing Round	No. of ER figs	No. that failed	% that passed
1	95	7	92.6%
2	93	2	97.8%
3	93		100.0%

The CR testing was carried out on a paper-by-paper basis. If one CR figure failed, the paper failed. The results are in the following table.

Testing Round No. of CR figs No. of papers No. that failed % that passed

1 105 22 10 54.5%

2 105 22 2 90.9%

3 105 22 100.0%

Testing Round	No. of CR figs	No. of papers	No. that failed	% that passed
1	105	22	10	54.5%
2	105	22	2	90.9%
3	105	22		100.0%

The NR figures were examined by the editors of each paper to determine if the figure should be considered non-reproducible or if the author needed to try to make it reproducible. There were 48 NR figures before the reproducibility testing for the ER and CR figures began and 50 NR figures after the testing was complete.

Next: Conclusions Up: Prucha, etal.: Reproducibility Previous: Testing reproducibility

Stanford Exploration Project
10/25/1999