Why the same figure appears differently in stories
Why one number can be reported in different ways across stories, through changes in definitions, start dates, denominators and rounding, and how to check each.
Two news stories can report the same underlying figure and leave readers with opposite impressions. One says a number rose sharply, and the other says it barely moved. Both may be accurate, because a figure is only one part of the message. The definition of what was counted, the dates used for comparison, the base the number is divided by and the rounding all shape how the figure lands. Reading it well means identifying which of these choices a writer made.
Before concluding that a story is misleading, check the source it drew on. Many differences come from an honest disagreement about a definition, or from a reporter picking a different reference point, rather than from an attempt to mislead.
Definitions and what gets counted
A count depends on what it includes. A figure for people experiencing homelessness may count only those sleeping in shelters or outdoors, or it may also include people in temporary accommodation. An unemployment rate follows a specific definition that differs from a count of people who filed for benefits. A figure for crime can mean reported offenses, recorded offenses or convictions, and each category can move differently over time because reporting habits change.
When two stories give different numbers, look for the definition in the original source. Reporters often copy a figure without its definition, so the story may say only "unemployed" or "crime." If the definition is missing, the figure is hard to compare with others. A change in definition can produce a jump that has nothing to do with the world the number describes, and a careful writer mentions such a change when it happens.
Dates, starting points and rounding
The choice of dates can change a story as much as the choice of numbers. A figure described as up 15 percent depends on which earlier period serves as the base. Another story might compare the latest month with the previous month, which can show a larger swing because of seasonal patterns. Neither is wrong, but they answer different questions. When a story gives a change, check whether the comparison is year over year, month over month or from an earlier peak, and whether the series has been seasonally adjusted.
Timing adds one more complication. Early figures are often revised as late data arrive, so two stories written a few weeks apart can quote different versions of the same series. A figure given without a date may be an early estimate that was later changed. When two numbers differ slightly, check whether one is a first estimate and the other a revised one, since the gap may come down to when each story was written.
Rounding adds another layer. A rise from 4.4 to 4.6 rounds to a move from 4 to 5, which looks like a full point, while the true change is two tenths of a point. Rounding rarely changes a story dramatically on its own, but it matters when the underlying change is small, and it matters more when different outlets round at different steps. If a small change is being presented as a big one, return to the unrounded figure in the source.
Denominators and how the number is framed
A count means little until you know what it is divided by. The number of incidents in a town of 10,000 residents and in a city of a million cannot be compared directly, but rates per 1,000 or per 100,000 can. Stories sometimes report totals when rates would be fairer, or report rates for a small group when totals would be more informative. Either choice can make a change look bigger or smaller than it is. A total can rise while the rate falls, if the population grows faster than the count.
The base can also change within a single story. A figure may be given as a share of all households, of households with children, or of people who answered a particular question. Each base answers a different question, and a good story says which one it means. A reader who misses the switch can take a change within a subgroup for a change in the whole population.
Money has its own version of this problem. A dollar figure may be given in the currency of its time or adjusted for inflation. Comparing a sum from one period with a sum from a later period, without adjusting, can overstate or understate the change.
Phrasing is the last lever. "Nearly half" and "just under half" describe the same share and leave different impressions. So do "one in four" and "25 percent." Word choice does not change the underlying data, but it shapes how readers feel about it, which is why the same number can appear alarming in one story and reassuring in another.
When two stories disagree, work through four points in order: the definition, the dates, the base and the rounding. Suppose one story says a count rose 15 percent and another says it rose 4 percent. If the first counts reported offenses and the second counts convictions, or the first compares with last year and the second with a peak two years earlier, the gap is explained before you reach the data. If the stories reconcile once those choices are visible, the remaining difference is one of framing. If they still do not reconcile, go to the original source and settle the question there.