I wrote:
> He has pen recorder data. he should publish it. I would suggest > normalizing it and putting it on top of the computer data in a single graph. > That would also be diddling with data. In a good way! You label the graph, "Pen recorder data normalized to fit the computer curve" -- or something like that. As long as you inform the reader what you are up to, you can do anything you like. If Parkhomov had published the first graph with an accompanying statement, "some missing data was replaced with random numbers" that would be okay. Weird, but okay. (It would be best to show the random number segments in a different color, or to show the original version of the graph.) This is partly a matter of standards. It is a matter of expectations in a given field. Nearly all of the ancient masterpiece paintings you see in major museums have been extensively repaired and painted over. If they hung the originals, you would see areas of missing paint, torn fabric, scrapes, missing faces and so on. What you see instead, with glistening paint that looks like it was finished last year, might be considered a modern facsimile rather than an actual painting by Rembrandt. It might be labeled: "Rembrandt and 20th century school of Rembrandt curators." The thing is, people who know about museums and curators all know this. It is not an attempt to deceive. - Jed
