Don't say you found aliens (unless you actually have)

Unlike with health and medicine press releases (Woloshin and Schwartz have a few good papers about the matter) I haven't seen much research about other scientific press release. That's why I was glad to find the paper "Credibility of science communication: An exploratory study of astronomy press releases" by Nielsen et al. (2007).

They conducted 11 in-depth interviews with journalists, scientists and public information officers, and came up with several conclusions regarding the accuracy and credibility of astronomy press releases.

• Credibility was defined by the interviewees as "being honest and doing your homework." Hype was defined as overstating the importance of results in order to increase visibility.

Credibility problems with press releases

Problems were usually caused by either the press release trying to make the issuing institution look better, or trying to make other institutions look worse.

1. The level of communication effort: finding some well-known person to tell the media how much the research described in the press release is significant to science.
2. The wording of the press release: even if it is possible you found alien life, keep that question mark at the end of "Alien life found?".
3. Dictating the timing of a press release: we just happened to discover something really important, just in time for the annual budget meeting! Other time-related sins are publishing the press release before the peer-review paper is out, and timing the release to screw up the competition's own press release or event.
4. Omission of reference to other scientists' work: this isn't the 17th century. It is hardly likely you did everything by yourself, or haven't built on some previous research.
5. Unjust comparison with other facilities.

Good ways to avoid lack of credibility (which can be in science, unlike in politics, problematic) is to have internal referees to the press release before its publication. Also, the importance of a peer-reviewed paper backing the press release can't be overstated.

Despited everything said here, the authors' overall conclusion is that "...credibility problems for astronomy press release do not exist, though examples certainly exist."

I find this conclusion very encouraging.

Nielsen, L. H., Torpe Jørgensen, N., Jantzen, K., & Christensen, L. L. (2007). Credibility of science communication: An exploratory study of astronomy press releases Proceedings from the IAU/National Observatory of Athens/ESA/ESO Conference, Athens, Greece.