Despite widespread recent news coverage of such mistakes as erroneously reporting Sinbad's death, people seem more likely than not to consider Wikipedia reliable, based on a recent poll conducted in the United States.
This information was part of a report produced by U.S. polling firm Rasmussen Reports. It addressed questions of reliability, how many people actually edit, and how favorably they view the project. The report was based on a survey of 1,000 adults conducted on 24 March and 25 March, with a margin of error of ±4 percentage points.
On the question of reliability, 46 percent found Wikipedia "at least somewhat reliable", with 16% disagreeing and the rest not sure. One-quarter of people who have visited the site said they had found something they knew was wrong, but only 9% indicated they had actually edited Wikipedia. The last figure corresponds roughly to the rule of thumb for participatory internet communities that about one in ten people will contribute instead of simply observing, and of those contributors about one in ten will be highly active. This type of "participation inequality" was studied with respect to Wikipedia last year by Jakob Nielsen and called the 90-9-1 Rule.
The percentage of people who had a favorable opinion of Wikipedia was the same as the percentage who found it reliable, 46 percent. Breaking things down by demographics, men under age 40 were especially well-disposed to the project, with a 68% favorable rating. Meanwhile, 20 percent of respondents had an unfavorable opinion of Wikipedia. Commenting on this, the report noted that based on some other recent Rasmussen surveys, "Wikipedia is not as well received as companies like Walmart and major auto manufacturers." (Interestingly, debates over the reception given Wal-Mart in its Wikipedia article have themselves received some attention in the past.)
While it is true that all four companies mentioned (Wal-Mart and the "Big Three" U.S. automakers, GM, Ford, and Chrysler) had higher favorable ratings than Wikipedia, all of them also had higher unfavorable ratings. In terms of favorable/unfavorable ratios, some did better than Wikipedia and some did worse. The only generalization this clearly supports is that each of these companies is better-known than Wikipedia, for good or ill. And in fact, it appears more than a third of people still don't know enough about Wikipedia to have an opinion about it.