As news writers, we must be careful to distinguish between fact and opinion. When we offer our own opinion, it should be clearly identified as such. When we offer someone else's opinion, we must say whose opinion it is. That, of course, means that we have to know to whom the opinion belongs in the first place.
One of the most obvious and most widely used non-attributable attributions used in everyday speech is "they say." In a news article that means absolutely nothing. The sentence "they say that the economy is getting better," gives the reader no information at all unless you have clearly identified who the word "they" represents. For example, "I asked three professors of economics from the Harvard School of Business, and they say..." identifies the people whose opinion or expertise is being cited and is a perfectly acceptable phrasing, especially if the names of the individuals are included or at least available via a source link.
The word "they" is pretty nebulous and is an easy one to catch, but sometimes, we might sound like we're giving source information when we're not. "Four out of five dentists agree that for patients who chew gum, sugarless gum is better." That statement is somewhat misleading because it seems to give the source of some bit of information, but does it? Did the writer only ask five dentists? That might make it a statistically unreliable sample. Did the writer ask twenty-six dentists until he got four within a five dentist sample that agreed with the premise? Did the writer ask 2000 dentists and get agreement from 1600 of them? Unless the detail of the survey is given, the reader doesn't know how much weight or credibility lies behind the statement.
As independent news writers, we need to work harder than someone who writes under the mantle of traditional news services like Reuters or the Associated Press to establish and maintain our credibility. So when a Reuters news writer uses the phrase "Economists say job growth between 250,000 and 300,000 is needed..." hardly anyone bats an eye or asks, "which economists? All of them? was it the economist cited in the previous paragraph and some unspecified others? Is it four out of five economists asked?" Strictly speaking, we, as readers, don't know whether that group of economists is employed by a biased entity, have any particular credentials, or even if they exist in any way more substantially than some guys in business suits overheard in a Starbucks near Wall Street.We need to be better than that. News writers should always give the reader the source of that information. If it is the consensus ascertained from a reputable poll or survey of economists, then we need to properly cite the polling organization and even the date of the poll, preferably with a link back to the actual data.
Here's an example of a vague source reference that is acceptable, because the specific information is given later on: "A government report says..." By itself that would not pass muster because it is unverifiable. There are too many government agencies and reports to allow the reader to judge the quality of the information provided. While I recommend providing the best information at the first use of the source's information, in this case the Associated Press writer does fill in the blank in the next paragraph when he says "The Department of Transportation's inspector general says in a report posted online Friday..."
That qualification gives the reader enough information to make a judgement about the information's credibility. Ideally, in the internet age, he would have included a link to the original report so the reader has access to the full set of information and the context from which the quotes were drawn. This wasn't possible when newspapers were the primary news outlet, but it is now and we should take advantage of linking to source material to enhance our news reports whenever possible, and not just when it is required in the citation guidelines of a particular news publisher.
In short, for any information which we attribute to the opinion or expertise any group of experts, we must say who those experts are and how we know the group believes it in order for our work to be credible. That applies whether the group is they, previous generations, veterans, Republicans, liberals, economists, or four out of five dentists.