Other People's Data

Taking other people's data and ideas and building upon them is what makes the sciences work. Unlike some areas of human endeavour, using other people's data is generally seen as a positive thing which benefits both parties. Just look at the academic success of the Sloan Digital Sky Survey team or the massively cited WMAP team who have provided easy access to their excellent catalogues. The reason the community are happy with this derivative work is because the original author benefits through a reputation built on citation; the people using the data will reference the origin and that helps the academic career of the original author.

The norms of scientific life mean that this system works, most of the time, but there are cases on the edges of what is deemed acceptable or ethical. This week Andrew Jaffe blogged about stealing scientific data with reference to a case where people had taken a photo of a slide of unpublished data during a conference talk and then extracted the data from the photograph. This sounds like a case of industrial espionage but it turns out that they did have permission from the original experiment's spokesperson so isn't as it first appeared. As Chris Lintott points out, putting your data in conference proceedings, in the arXiv, or having people discuss them is a form of publication. Although not peer reviewed, the buzz surrounding the unpublished data could well mean that they receive plenty of citations once they are published formally.

There are some other cases where the people making use of other people's data do not cite the original work and claim the credit for themselves. One example of this in the past few years is a case of spying on somebody else's observing logs and scooping them to the discovery of a Kuiper Belt object. That sort of behaviour is underhand and unfair on those putting in all the effort to acquire and process the original data. Although these cases are probably quite rare, I am becoming more aware of the issues.

One of my astronomical friends has been working extremely hard for a couple of years on a huge survey of the galactic plane. His team have taken hundreds of hours of observations, using a big telescope, specifically for this project. As is usual on big telescopes, they had an agreed proprietary period with which they could analyse the gigabytes of data before they became free to all. This proprietary period lets those who put effort into getting the observations made exploit them and publish their results first. The period can be anything from 6 months up to about 2 years depending on the telescope. Unfortunately, in my friend's case the data were accidentally released early and without warning. A rival group took the data, did a fairly quick and dirty analysis of it and then published them as their own.

As you may imagine, my friend isn't happy. He has agonised long and hard over those data to tease out as many systematic uncertainties and calibration errors as he can. The rival team have just done a fairly quick analysis and then taken all the credit. Will this theft of credit from my friend mean a reduction in citations for his work when he publishes it. I really hope not but if it does, that could hurt his career. I don't think my friend's case is typical but it does serve as a reminder of the impacts of unscrupulous researchers on other's lives.

Is there a moral to this? Yes. Always give credit to those whose work you use. To give them credit doesn't cost anything, it benefits them and is the right thing to do.

Tags: | |
Posted in astro blog by Stuart on Thursday 04th Sep 2008 (13:45 BST) | Permalink
[an error occurred while processing this directive]
[an error occurred while processing this directive]