You may have seen this article in Salon on “American men’s hidden crisis: They need more friends!” Slate picked up on the same theme as well. It is a juicy article, but I got tripped up on the first sentence: “Of all people in America, adult, white, heterosexual men have the fewest friends.” (Slate’s version is: “… adult white heterosexual men have fewer friends than any other group.”)
The Salon article is hyperlinked to a source – Social Isolation in America: Changes in Core Discussion Networks Over Two Decades. It is a thoughtful paper, grappling with some novel survey data and the social isolation/loss of social capital work that started swirling around in the 1990s (think, Bowling Alone).
The problem for Salon (and by extension Slate) is that this scholarly report doesn’t say “white heterosexual men have fewer friends than any other group.” It says white men have the MOST friends.
Seriously. Look at the chart:
The chart shows that women were lagging behind men significantly in the number of non-kin members of their confidant network (what we call “friends”) in 1985, but by 2004, while women had fewer friends than men, the deficit was no longer significant. African Americans and those of other races lagged behind whites significantly in the number of non-kin confidants in their network.
It can be a little confusing – women tend to have larger overall networks because they have more regular contact with family.
But even if you were confusing “friends” with “family” in these networks, how do you miss statements like this: “Non-whites still have smaller networks than whites?”
And nowhere in the report is sexual preference even mentioned, so how can this be a source for the “heterosexual” descriptor in this claim?
Here’s why this is not a good thing – the writer (in this case a sociologist and faculty member at Occidental College) is misrepresenting a source. This is not about whether the writer’s overall argument is correct, and for all we know the writer might have done a little more research and found a report that actually said something like “Of all people in America, adult, white, heterosexual men have the fewest friends.”
But the given source appears to contradict part of what the writer claims, and doesn’t address at all other parts of that claim. And whatever the cause of the error – a momentary lapse in reading comprehension, rushed or sloppy editing, the power of narrative to make facts seem bothersome – the larger point is that this article was presented as fact-based inquiry in a reputable journalistic outlet.
The meaning of a hyperlink inserted in a claim you make in an online article is clear: “you can trust this statement. It is based on good evidence. Accept the statement and move on with the article.” In this case, though, that isn’t true.
So, if you don’t really have a source for your statement, don’t just link to an article that sounds kind-of like the evidence you would like to have.