Home/Research/Three Is a Magic Number

Counting Sources

One measure of evaluating objective stories

By Bradley Wilson and Lee Williams
June 1, 2007


Textbooks frequently include a number of criteria for evaluating news stories, everything from the form of the lead to the order of quotations and transitions. While numerous texts refer to the ideal types of sources, few give any ballpark figure for an acceptable minimum number of sources. For this study, we didn’t set out to prove what the ideal number was, but we did attempt to determine what the actual number of sources being used was in a case study of a daily student newspaper at a large, state university in North Carolina. We found that the average number of sources in objective stories (news and features) was 2.83, with an average of 1.60 students used as sources on average per article.

In a related study about four years ago, a professor at Kansas State University found the average number of sources in college newspaper stories was 3.54. Seems reasonable. For years, educators have been teaching that the ideal number of sources in an objective story was at least three: one to tell one side of the story, one to tell the other, and one to break the tie.

But, at the time, a cursory analysis of our student newspaper, a 10,500 daily circulation, student-run newspaper at a land-grant institution with no journalism degree program, showed we were well below that. So, during the past four years, we’ve tracked the number of sources in each and every article in the paper in an attempt to determine the actual number of sources our reporters used in comparison to the 3.54 benchmark.

Now, to be fair, at this point, we would include a review of what past literature had shown. But despite repeated searches through online materials and textbooks, and direct communication with a variety of scholars, we found almost no material regarding the ideal number of sources. While there was a plethora of material on the ideal types of sources, even guidance on how to seek them out, there was none on the ideal number of sources. So, we interviewed educators and professional journalists to seek out their advice, also in an admittedly unscientific, but informative, manner.


When it comes to future research, a more scientific poll might get at what educators and journalists believe is the ideal number of sources. However, the results are likely to be similar to the informal results we found from both academics and newsroom professionals: “It depends on the type of the story and the circumstances.”

For example, Mike Haynes, at Amarillo College said in classes he requires students to include at least three sources, and encourages students in second-semester classes to have five. He said he believes that the student publications probably have about four sources each in their stories. However, he also conceded that, “I don’t like to enforce a ‘source quota,’ because some stories need several sources, and some don’t. But for beginning writers, our three-source minimum works well.”

Bill Neville also said three sources was the minimum number of sources he desired when critiquing the paper, but “the more the better and, perhaps, closer to the truth.” Joe Gisdondi, adviser of the Daily Eastern News at Eastern Illinois, and an assistant professor of journalism, also vouched for the magic number three, adding that it might hinge on the type of sources. “I require three sources regardless if the story is a precede or a comedian on campus. … Too many reporters, though, use this as a magic number, grabbing the first three people they can find. Instead, they should spend time with sources, perfecting the ‘fine art of hanging out,’ as someone called Gay Talese’s approach.”

It was Mark Plenke of Normandale Community College, who called into question the validity of the magic number three. “Professional reporters and editors know there is no magic number of sources for stories to be successful. If a reporter is writing a 3-inch brief about an upcoming event, it’s probable one source will do. It sounds as if evidence of contacting three sources needs to be evident in the stories you’re grading, so the policy invites longer-than-necessary stories — a poor thing to teach. On the other end of the scale, a story about a riot on campus should probably have dozens of sources. Three would never be enough.”

Like the academics, editors and reporters in central North Carolina responded with varying opinions regarding the appropriate number of sources in an objective story. While some suggested that three to four sources might be an appropriate average for a news story, others admitted that attempting to arrive at some sort of “magic number” for sources is a difficult task. These professionals were nearly unanimous in their assertion that the number of sources in a story is almost always contingent on the type of story and the circumstances surrounding it.

Wendy Lemus, editor of the Cary News, said while she viewed three or four sources as average for a story, ultimately “it really varies depending on the story.” And working for another McClatchy publication, Josh Shaffer, news reporter for the Raleigh News & Observer, said, “for a standard, 15 to 20-inch daily story, I’d say a minimum of three. I’d be embarrassed to hand in a story with only two sources, but it happens sometimes.” Shaffer added that for longer, more intricate stories, a minimum of five sources is an appropriate number, specifying, “five quoted sources,” with other archived materials or reports possibly thrown in for support. “The more sources the better,” Shaffer added.

News & Observer Executive Editor Melanie Sill said while few stories can be written based on an interview with a single source, some stories contain sources that are not directly quoted and are used as supplementary sources. “Often you won’t recognize from reading the story that there are multiple sources,” Sill said. “A reporter might talk with people or consult written sources for background or clarification, but not quote those sources in the article.” Sill added that, “a complex story may have dozens of sources in terms of people, documents or other reference material used in research.”


Case studies aren’t without their problems, not the least of which is lack of generalizability. This case study is not representative of the entire population, nor does it claim to be. Researchers using the case-study method can only generalize what they learn from one specific case to cases similar to that one.
Jensen and Rodgers (2001) identified five types of case studies.

  1. Snapshot case studies: Description of a single organization (or other entity) at a single point in time; the least meaningful according to critics but have a distinguished history in public administration.
  2. Longitudinal case studies: A time-ordered analysis of events that occurring during a period of the entity’s history.
  3. Pre-post case studies. A longitudinal-type design but includes an assessment before implementation of some program, policy or decision and follow-up assessment after implementation; causal inferences can be drawn.
  4. Patchwork case studies: Integration of several case studies that have evaluated a particular entity at different points in time; creates a more complex and holistic picture of the dynamics associated with the entity.
  5. Comparative case studies: cross-unit comparison as opposed to within-unit synthesis; an attempt to tease out generalizations about underlying commonality reflecting a policy, process, program or decision.
  6. In this longitudinal case study, we did not set out to determine what the ideal number of sources was, we did set out to determine how many sources our student reporters used on average over time, to give us a baseline against which to measure other stories. It also became a measure of the success of various editors as we discussed the importance using multiple sources to help the paper be accurate and fair. In all, over the last seven semesters (including two summers), we found the students edited 4,052 objective news/feature articles (not sports or opinion/editorial) containing 11,470 (not unique) sources for an average of 2.83 sources (and 1.60 student sources) per story, significantly lower than the “desired” 3.00 and well below 3.54. (CHART 1)

But measuring the number of sources used in each objective article wasn’t the only thing we tabulated each day over nearly 400 issues of the publication. We also tracked how often the staff met deadline, and rewarded the editor if the staff met deadline more than four days per week (>80 percent of the time). We tracked the average number of student sources used in articles, the percentage of articles that did not use any apparent sources (i.e. were reprinted from press releases or other media without citing a source) and the percentage of articles in the paper that used more than three sources. (TABLE 1)

While it started as a way of performing a check and balance on the business office and getting an approximation of the income based on advertising, we also started tracking the percentage of the paper devoted to objective news and feature stories, sports stories, opinion/editorial, advertising and the average number of pages. (TABLE 2)

At the end of every month, we prepared a report highlighting significant changes, progress towards our goals and items of concern, comparing the current month to that semester and past years. For example, over the seven semesters studied to date, the average number of sources in the paper has ranged from 0.71, on a Monday right after spring break when we had a high story count but few sources, to 5.58 on a Wednesday right after the election in the fall of 2006. The average number of students used as sources in that 10-page paper was proportionally high at 3.92.

Just for comparison, at various points during the year, students themselves would analyze a newspaper from another school so they could understand the methodology being used to evaluate their paper. Although chosen at random, the newpapers, including single copies of some papers and multiple copies of other college newspapers such as The Daily Texan at the University of Texas at Austin, The Daily Tar Heel at the University of North Carolina at Chapel Hill and The Ithacan from Ithaca (New York) College. The results were similar. With an average of 13.6 pages and 17.4 stories per issue, these 64 issues of 28 different papers had an average of 2.65 sources per story and 1.67 student sources per story, validating the data we found in our own paper.


The first thing that became clear as part of our analysis and discussions during the past four years is that having three sources is a good goal for most stories. Three promotes at least some balance. And it gives less-experienced reporters something to shoot for.

The second thing that became clear was that the “right” number of sources completely depends on at least two other factors: the type of story and the type of sources used. A news brief about an upcoming club activity may service the campus with few sources. A story about a controversial new policy or ineptness of some administrator may require many sources to paint the full picture in a fair manner.

Several newsroom professionals validated this point. For example, Bob Ashley, editor of The (Durham) Herald-Sun, attested to the contingency of number of sources on the type of story and sources. “There are some simple and straightforward stories in which a single source is quite adequate. There are more complex stories in which one might talk to multiple sources and still wish we had time to talk to more.” And Dan Holly, editor of the North Raleigh News, pointed out different types of stories that would require different numbers of sources. Holly referenced a human-interest story about an accident victim who overcame serious injury and earned a black belt.

This was, according to Holly, an appropriate one-source story because “the chances of an elaborate deception were slim.” However, other types of stories such as disputes or allegations of wrongdoing need at least two sources “who are in a position to know have no axe to grind.”

Some professionals such as News & Observer reporter Sam LaGrone question even trying to propose a universal minimum for news stories. “It’s akin to asking an engineer how much steel he or she needs to build a bridge…Some take four. Some take hundreds.”

So, like all case studies, this case study doesn’t complete the picture. It is but a look over seven semesters of one college daily newspaper. It is not a measure of what is right in terms of the number of sources, it is measure of what is. While in an unscientific comparison to other college newspapers (dailies and non-dailies), this newspaper does seem comparable to other college papers. The next step in this research is for other papers, professional, college and even scholastic, to calculate the average number of sources their reporters use in objective articles. But even that can’t be the end.

The professional journalists and educators have all raised valid questions and thrown in a bit of skepticism when it comes to the search for an ideal number of sources. So, the number can’t be used as an ideal standard, but can be used as an objective comparison. To carry this research to the next logical step, industry professionals and educators should break down the research to determine the average number of sources used in various types of stories, everything from breaking news to investigative stories (which we suppose will have a greater number of sources) to routine feature stories and even personality profiles (which we suppose will have fewer sources).

Clearly, the “ideal” number of sources is like finding the ideal number of pages for a paper in English class, “enough” but “not too many.” In the meantime, we’ll have to stick with the magic number three sources per article as a goal, knowing that on average, college papers seem to use slightly fewer than that.

APPENDIX: Semester-by-semester charts


  • Ashley, Bob. “Re: question about sources.” Email to Lee Williams. 4 May 2007.
  • Gisondi, Joe. “Re: sources, informal comments.” Email to Bradley Wilson. 8 May 2007.
  • Haynes, Mike. “Re: sources, informal comments.” Email to Bradley Wilson. 8 May 2007.
  • Holly, Dan. “Re: question about sources.” Email to Lee Williams. 4 May 2007.
  • Jensen, J. L., & Rodgers, R. (2001). Cumulating the Intellectual Gold of Case Study Research. Public Administration Review, 61(2), 235-246.
  • LaGrone, Sam. “Re: question about sources.” Email to Lee Williams. 20 May 2007.
  • Lemus, Wendy. “Re: question about sources.” Email to Lee Williams. 8 May 2007.
  • Neville, Bill. “Re: sources.” Email to Bradley Wilson. 7 May 2007.
  • Plenke, Mark. “Re: sources, informal comments.” Email to Bradley Wilson. 8 May 2007.
  • Shaffer, Josh. “Re: question about sources.” Email to Lee Williams. 18 May 2007.
  • Sill, Melanie. “Re: question about sources.” Email to Lee Williams. 7 May 2007.
  • Simon, Todd. “Re: content analysis.” Email to Bradley Wilson. 22 May 2007.


  • Gibbs, Cheryl, and Tom Warhover. Getting the Whole Story: Reporting and Writing the News. Guilford: New York.. 2002.
  • Harris, Julian, Kelly Leiter, and Stanley Johnson. The Complete Reporter: Fundamentals of News Gathering, Writing, and Editing. 4th ed. Macmillan: New York. 1981.
  • Harrower, Tim. Inside Reporting: A Practical Guide to the Craft of Journalism. McGraw-Hill. New York. 2007.
  • Jensen, J. L., & Rodgers, R. (2001). Cumulating the Intellectual Gold of Case Study Research. Public Administration Review, 61(2), 235-246.