Here in NY, subway posters from the transit service have a warning picture of a figure leaning far over the platform, with the title “don’t become a statistic,” followed by smaller text on the high number of fatalities in 2013 from falling on the tracks, etc.
We all understand that, of course, and the use of a once arcane (?) science-y sounding word is chosen for a reason. Instead of saying “don’t wind up like 534 other people who were hit by trains last year”–“a statistic” conjures up the sadness and strangeness of what was once a human life reduced to a number used by other people for reasons that have nothing to do with amyone’s, especially your lived life. (In a follow-up post I’ll cite a Heinrich Böll short story bases on this.)
Got it. The average subway rider is expected to know and respond to the word “statistics” as generally understood in modern society.
Again, I wonder when the word/concept spread to general vocabulary. Maybe broadcast election coverage, with reporters using (usually improperly understood) a new tech-sounding analytical word, like how they use “social media response.” Or how they constantly misunderstand statistical results, come to think of it.
But as a matter of strict usage, I don’t think “a statistic” makes any sense. A “data point” is how I would say it, but maybe professionals in the field have a different, more accurate term.
Any ideas on the definition or the history of common knowledge of the field?
You’re right, that is more accurate usage. As someone who has taught statistics to undergraduate poli sci students and who uses statistical methods in his work, my take on ‘common knowledge of the field’ is that there is an arresting degree of statistical illiteracy which prevents people from producing and consuming quantitative information properly.
“A statistic”, singular, is a well-defined mathematical concept. It’s just a measurement of some data set of interest, or it can refer to the function that produces that measurement. The mean number of people who die in subway accidents each year is “a statistic”.
Now, that mathematical definition doesn’t map well to the colloquial definition. To be pedantic, the subway platform safety poster should say “don’t become part of a statistic” or “don’t become just a data point” as you say. But you could also use the mathematical definition to produce “a statistic” that corresponds to a single person or event; just use a function that takes a data set and outputs the nth data point. I can’t see how that statistic would be useful, but it might be in some case that I’m not aware of.
A good read is I Bernard Cohen’s The Triumph of Numbers. More history of statistics, but it has a reasonable overlap with the question. I’ll hunt out my copy later.
Statistics does have (at least) two meanings. One is simply the gathered data. The other is the study of data, and especially the elicitation of meaning from that data. A single point of data is reasonably called a statistic IMHO.
I think “become a statistic” is an idiomatic phrase, and the average* subway rider is expected to be familiar with that idiom.
*(Note that “average” is itself a statistical concept that is being used imprecisely and unscientifically here.)
Even if more technically correct, “data point” doesn’t seem to scan, or convey detached bureaucratic coldness, as well as “statistic” does in the Police’s “Invisible Sun”: I don’t ever want to play the part
of a statistic on a government chart
(Of course that song immediately entered my brain after reading the OP…)
I tend agree with the claim that the word “statistic”, singular, as questioned in the OP, is a colloquial usage meaning “an individual data point” and especially a tragic one (as in “don’t drink & drive & become a statistic”)
The term “statistics” derives from “statist” or “statism”, which derives from “state”, and refers to the collection and use of demographic data by governments for driving public policies.
Wikipedia article has a section on the history of statistics, suggesting it goes back, in some form, to 5th century B.C. Statistics in its modern form began to develop in the 16th century, and in particular with the development of probability theory by Pascal and Fermat in the 17th century, with much additional theory developed from the 19th century to present.
The difference as I see it is that a “data point” is rather neutral. We are all data points already. Everything than can be measured about us is a data point. My current data point is “has not fallen onto railroad tracks”.
“Becoming a statistic” kind of implies becoming part of a group, a specific set of data points. In this case, the group is those who have fallen onto tracks.
I’ve heard it all my life, thanks for confirming it is an old usage. This is the message I take from the usage - don’t do something to convert yourself from an individual with a name and a future into a statistic in the list of the dead.
Well its confusing to say “data point” is the same as “statistic” .
Often one mentions a “data point” to give a context for the discussion… to help keep the information real or clear… because sweeping generalizations can be too easy to make…
Well a statistic is used to mean “uninterpretable numbers” , or even " sweeping generalization" …
For example, what if you were writing an article about the Falkands War.You might describe how the Navy did this ,and the helicopter did that, and the men marched and the men shot , and 156 men died here and 235 men died there" and so on. Well thats statistics.
A documentary on the war may then pick some individuals , and tell of their involvement in the war, perhaps their training, preparations, and injuries and points of view… thats a clarifying data point… or data that isn’t statistical.