Explore Harvard's Nieman network Nieman Fellowships Nieman Lab Nieman Reports Nieman Storyboard

Lexis searching, like Google, has numeric pitfalls

SHOWCASE | October 298, 2005

Results can vary according to what type of subscription you have; unique hits may be far fewer than the total cited.


By Alex Kingsbury

akingsbury@niemanwatchdog.org

 

The Lexis-Nexis database may not enjoy the ubiquity that has made Google a verb, but it is nevertheless indispensable for many reporters and editors. Nexis, the news portion of the legal and media database, is used much the same as Google, as a tool for cross-referencing the collective memory of the media and harvesting information that can inform tomorrow's news.

 

Like all tools, however, Nexis searching, which has been available far longer than Google, can be misleading.

 

The dangers inherent in citing Google search results are well documented. The search engine's method of collecting and tabulating results can lead to over or under-reporting, based on the way the query terms are entered. Citing Nexis results, on the other hand, continues unabated by journalists and pundits, though the practice suffers from similar structural flaws, the least of which is the fact that only a portion of the country's newspapers and printed media are included in the database.

 

Consider the following examples, retrieved, of course, through Nexis.

 

On August 28, 2005, the Boston Globe cited search results in a staff editorial. "(Karl) Rove's connection to the Valerie Plame story was the center of attention in mid-July but cooled fast after Bush nominated Roberts to the Supreme Court on July 19. A Lexis Nexis search reveals 1,944 stories mentioning Rove in the week prior to the nomination, dropping to 1,111 during the week after."

 

For the sake of argument, let us query the name "Karl Rove" for July 19. It yields 222 hits, only a portion of those cited in the Globe editorial. A closer examination of these hits shows that of those 222 mentions of the term "Karl Rove", I found only 81 unique news *stories*. Thirty-two of those hits were letters to the editor and 59 were staff editorials.

 

Some hits were the same story published in different editions of the same newspaper. Many were multiple hits of nearly identical wire service copy with a slightly altered word count or headline. Ken Herman of Cox Newspapers, for example, earned six unique hits in this search because his story appeared in six different papers.

 

Citing Nexis is also problematic because the results can be difficult, if not impossible, to replicate depending on how much of the Nexis archive your subscription covers. On March 5, 2004, the Daily Standard wrote, "If you search the Lexis-Nexis database for mentions of (Ralph) Nader on February 22, 2000, the day after he announced he would run, you'll come up with only 73 hits, worldwide." Now, a search of the term "Ralph Nader" on February 22, 2000, returns 43 hits, worldwide.

 

The American Journalism Review regularly features Nexis search results in its "Cliché' Corner". "Whiff", as in 'whiff of a scandal' or 'whiff of panic,' appeared more than 750 times in February 2005, according to AJR. "Yeah, we get that number from Lexis-Nexis," says Managing Editor Lori Robertson. She acknowledges that the search results often contain multiple hits of the same story or transcript in the Nexis archive. "We do a search and make a rough estimate to give readers an idea of how many appearances a particular cliché has made."

 

While the camera might never lie, sometimes the archives do. In 2001, Forbes magazine noted that 10 major US newspapers erroneously reported that Mexico's newly elected president, Vincente Fox, earned a degree from Harvard University. Despite being widely reported, Fox did not attend or earn a credential from that school. "I call this phenomenon the 'Nexis virus,'" says Sree Sreenivasan, dean of students and professor of new media at Columbia University's Graduate School of Journalism. "Journalists tend to assume that a fact is correct if it has been printed in ten major publications. Obviously, this is not always the case."

 

On September 30, 2005, the Calgary Herald reported that President Fox "studied business administration and management at Mexico City 's Jesuit-run Ibero-American University and at Harvard." The Calgary Sun had a story on the "Harvard-educated Fox" a two days later.

 

And on May 22, 2005, despite being cited four years earlier in the Forbes article, a member of the Chicago Tribune's editorial board mentioned "Fox and fellow Harvard Business School graduate President Bush" in an op-ed. Of even more concern, is the echo chamber effect that the database has had on journalism and the scope of the public debate. Nexis is often the first stop for reporters, who are then prone to calling the same sources quoted in other publications. "The same people are called over and over again because they are in the Rolodex of the gatekeepers," Sreenivasan says.