NASA has denied covering up its first survey of airline pilots on safety, despite releasing the figures late on New Year's Eve and in pdf format rather than Excel files, making independent number-crunching more difficult.
The National Aviation Operational Monitoring Service talked to 8,000 pilots between April 2001 and December 2004. It received more than 25,000 responses to its survey.
Early indications suggested that the research found that incidents like near-misses and runway problems occur much more often than previously believed.
But NASA's decision to redact the information and to only publish it in pdf format severely limited the media's ability to interpret the data, and sparked suggestions it was burying bad news.
Whole rows and columns of data have been removed in order to maintain anonymity, rows were also reordered and individual entries were selectively edited.
NASA originally delayed releasing any of the information because of fears it would damage the aviation industry. NASA spent $11m on the research.
NASA administrator Michael Griffin said the information was not of any interest to travellers. Griffin said the agency was not intending to do any more work with the data which was collected to test different methods of gathering data rather than to prove anything about airline safety.
The space agency all but disowned its own data. A disclaimer on its website warned that neither the survey responses nor the data had been peer-reviewed. Therefore NASA warned: "No product of the NAOMS project, including the survey methodology, the survey responses, and any analysis of the responses, should be viewed or considered at this stage as having been validated."
US politicians have promised to push NASA to release more information. The survey was originally to be extended to include airline mechanics and cabin staff.