"Obama’s Support Softens in Poll, Suggesting a Peak Has Passed" By Adam Nagourney and Megan Thee for the New York Times.
In an article Friday the New York Times reported Senator Barack Obama’s support among Democrats nationally has softened over the last month, particularly among men and upper-income voters, as voters have taken a slightly less positive view of him than they did after his burst of victories in February, according to the latest New York Times/CBS News poll. The article used alot of numbers from the poll data to tell the story in a straight forward an direct fashion.
An example paragraph from the article:
"Mr. Obama’s lead among men has disappeared during that period. In February, 67 percent of men wanted the party to nominate him compared with 28 percent for Mrs. Clinton. Now 47 percent back him, compared with 42 percent for her, a difference within the poll’s margin of error. Similarly, his lead has shrunk among whites, voters making more than $50,000 annually and voters under age 45."
Multiple numbers are used within this paragraph but they all support a similar context and work to tell a clear story. In my opinion the use of numbers is easy to grasp, however I could see how some would have difficulty. It may have been beneficial to break the information into seperate paragraphs and set a maximum of numbers used in a paragraph at two.
For example the paragraph could instead read, "Obama's lead among men has dissapeared during that period. Since February Obama's has lost men in his party's support by 20 percentage points, while Clinton has gained almost as much in support.
Now 47 percent back him, compared with 42 percent for her, a difference within the poll’s margin of error.
Similarly, his lead has shrunk among whites, voters making more than $50,000 annually and voters under age 45."
Overall, the numbers are used effectively, and without the exact results the reporters were given it is unclear how much number crunching was involved, but it appears their was good math reporting being done.
All numbers appear to be taken from the poll. No outside sources appear to have been used.