NEW YORK – The Internet ranks behind only television as the leading source for science news and information, but most users won't trust what they read online blindly, a new study finds.
The Pew Internet and American Life Project said in a report Monday that 20 percent of Americans obtain most of their science information from the Internet, compared with 41 percent who cited television.
Newspapers and magazines were each credited by 14 percent, and radio by 4 percent.
• Click here to visit FOXNews.com's Home Computing Center.
The gap disappears among users of high-speed Internet connections at home, with 34 percent saying they turn to the Internet most of the time, and 33 percent citing television.
About 80 percent of those who get science information online try to check its accuracy elsewhere — another online source, offline resources or the original study — and many of them use more than one alternative.
Only 13 percent say they turn to the Internet primarily for its accuracy. Most do so because they consider it convenient.
Americans, however, rely more on the Internet for science news than general news, Pew found.
While the Internet ranked second behind television for science news, it was behind local and national television, radio and the local paper as the typical source for general news. It beat only national newspapers. (The questions in the science survey did not split TV and newspapers into local and national.)
"There's a lot of good scientific content out there," said John Horrigan, Pew's associate director. The Internet was "initially about science and engineers talking to each other. That community has a historical head start in terms of getting information online that's useful to science consumers."
Some 87 percent of Internet users have looked up science information online at one point or another, and two-thirds say they have stumbled upon science news when they logged on for another reason.
The study was based on telephone surveys of 1,447 Internet users conducted Jan. 9-Feb. 6. It has a margin of sampling error of plus or minus 3 percentage points.