Numbers Don't Add Up at New Star-Advertiser
Three Sundays. Three lead stories in the Honolulu Star-Advertiser based on numbers. Three times the figures are misused and don't add up. Once, OK. Twice, a concern. Three times, I'd say that's a problem.
The perception I come away with is that the paper is trying to make a splash every Sunday. Look at this, it shouts. Can you believe how bad this is? But when you look closely, its front-page headlines aren't supported by the numbers.
Here are the lead headlines from the Star-Advertiser on its first three Sundays:
- "Fresh Costs"
- "ON OAHU, CRIME UP" (Yes, the headline across the entire width of the front page was in all capital letters.)
- "Diploma distress"
So let's take a look at whether the headlines delivered what they promised.
Sunday, June 13: My article about the first Sunday story, The Weekend Editions Of The New Honolulu Star-Advertiser , explained that "Fresh Costs" referred to the fact that fruit and vegetable distributors in Hawaii are paying about $300 a week in unexpected expenses to bring in inspectors on overtime. The piece claimed that was an important factor in driving up our grocery costs. If you do the math, the extra costs over the six months the article covered were $7,800. That gets spread over a population of 1.3 million people, plus nearly a couple of hundred thousand tourists, and tens of millions of pounds of fruit and vegetables. Looked at that way, the number just isn't material.
Conclusion about the first Sunday: All heat, no light.
Sunday, June 20: "ON OAHU, CRIME UP" screamed the front-page headline. The secondary headline said, "Major offenses rise islandwide, prompting neighbors to unite in community watches and patrols." So did the story ever tell us what a major crime was? No. At least not explicitly. But what a graphic on the front page made clear was that violent crime had essentially stayed flat. The number was actually down 38, from 2,575 in 2008 to 2,537 in 2009. It also told us that the figure for violent crime for 2009, the year the article covered, was down 208 from 2006, the oldest data it presented. That's a decline in violent crime of 7.6 percent over four years, and the number was lower than the year before every year.
The only way a reader could figure out what "major offenses" meant was by surmising that the categories of burglary, larceny and auto theft for property crime and aggravated assault, robbery, and murder and rape for violent crime in a graphic on the front page were "major offenses." Maybe those are what make up major offenses. But the paper never made it clear.
The only evidence in the entire story to support the secondary headline that neighbors are uniting in community watches and patrols was a paragraph paraphrasing Meri Peter, coordinator for a neighborhood security watch in Wahiawa. It said: "More people appear to be participating in watches than in the past as they see crime increase in their neighborhoods." That's it. Come on. Since when does an assertion from one person make for a trend? If it had been true, it should have been easy for the Star-Advertiser to profile some of these concerned citizens, to make the impact of the "spike" in crime real for its readers.
Finally, buried at the end of the article based on the Honolulu Police Department's annual report released earlier in the month, a determined reader would have discovered that the number of people arrested in connection with major crimes rose 19 percent. Yes, 19 percent. Now that's a spike. And it might have actually been the real story.
The front page story on the second Sunday was an example of how numbers shouldn't be used. When you lump all major offenses together and focus on only one year, the instrument is too blunt. Violent crime is obviously the most important category. Personal safety comes first for people. But a reader wouldn't have a clue about the state of violent crime in their district from the Star-Advertiser's story or graphics.
Conclusion about the second Sunday: Another case of too much glitz in the graphics and too little thought in the content.
Sunday, June 27: "Diploma distress" was the headline. The secondary headline told readers that graduation rates are "slumping." So what did we learn from the numbers? By now you can probably guess. The headline wasn't supported by the paper's own data. The graduation rate in the 2008-2009 school year, the latest year available, was exactly the same as the year before: 79.9 percent. And it was higher than for every other year from 2001 on. The reality is that the rate has essentially stayed flat for a decade.
The beginning of the Star-Advertiser story said that eight public high schools among the state's largest campuses (whatever that means) have graduation rates under 75 percent. The problem is we never find out whether that's worse than previously or not. (At least we can find the names of the eight in the body of the story.) The second part of the story's first sentence tells us that 28 schools saw their graduation rates decline in 2009. That must be the slump the secondary headline was talking about. The problem is that when the paper produced a chart showing the graduation rates for all public high schools, it didn't include what the rate was the year before. So as a reader, I haven't a clue which schools saw their rate drop or by how much. The overall rate stayed flat, after all.
Conclusion about the third Sunday: This problem is starting to seem like a pattern. The Star-Advertiser's numbers aren't helpful.
I know the Star-Advertiser has decided the web isn't worth much attention, but it might reconsider when it comes to stories involving numbers. The web is the perfect place for it to put data that readers can use to better understand their community. Imagine if the paper had posted sortable charts where parents could have compared their school with other schools or looked at graduation rates through a number of filters, such as poverty or ethnicity. And working the numbers this way might help its journalists, too.
Now that Honolulu has only one newspaper — and the most widely distributed weekly is owned by the same company — it has a higher level of responsibility than either of the papers it replaced. We've yet to see its Sunday front page live up to the challenge. That has serious ramifications for the community's understanding of itself, not just for the journalists responsible for the work.
DISCUSSION: Do you have concerns about what it means for Honolulu to be a one-newspaper town? I'm going to talk about my observations and discuss the issue at a Beatup in our office at 5:30 p.m. on Thursday. If you want to attend, please RSVP to email@example.com so we know how many people to expect. And in the meantime, share your thoughts on the Star-Advertiser and the city having only one daily newspaper in our discussion on the topic.