They are profoundly unserious, innumerate and misleading.

From "And nearly 13 percent — 20 percent of the lesbian couples and 5 percent of the gay couples — participated in some version of 'undoing gender.'" by Ann Althouse.  Her headline is from a 2018 NYT article.  The substance of the article is minuscule.  However, it does provide another example of the astonishing innumeracy of journalists.

5. The NYT article has a correction noted: "An earlier version of this article misstated the percentage of parents in a study who were 'undoing gender' by creating new names. It is nearly 13 percent, not a quarter." I read the comments over there and saw how the mistake was made. As you see in the quote in my post title, they had "20 percent of the lesbian couples and 5 percent of the gay couples." They added 20% and 5% to get 25% ("a quarter"). That's some embarrassing innumeracy. 

Woof.  Doing a calculation and making a mistake is one thing.  But simply not understanding how percentages work is at the next level.   

And it is still not correct.  Adding 20 + 5 to get 25 and then dividing by 2 to get "nearly 13%" (actually 12.5%) only works if you know that there are an equal number of gays and lesbians.  But that is not the case.  

It is true that in the woefully ill-designed study on which the article is based is also woefully underpowered (80 people, 20 lesbian couples and 20 gay couples) who were not randomly selected and therefore not representative.  

If you really want to know what percentage of LGBT couples are "undoing gender" and you are collecting data by male single sex couples and female single sex couples, then you have to take into account two additional factors- the relative prevalence of those gay and lesbians couple in the population and then the respective disposition of lesbian couples and disposition of gay couples to raise children.  

Fortunately between Gallup, The Williams Institute and the American Community Survey, we do have that data.   According to the latest research by Gallup, 0.7% of the population is lesbian and 1.4% are gay.  

On the other hand, 8.1% of gay couples have children whereas 24% of lesbian couples have children.  

Gay couples are twice as common as lesbian couples but lesbian couples are three times more likely to have children than gay couples.  

But this is beyond the purview of the article since it is based on the poorly constructed research being done.  The New York Times chose to publish an article based on still peer unreviewed research with a facially underpowered sample size (forty couples instead of several hundred) of a non-randomly selected population.  Neither the research nor the article took into account the fact that gay couples are twice as prevalent as lesbian couples nor did either take into account that lesbian couples are more than three times as likely to raise children as gay couples.

This article should not have been written at all, given the shamefully inadequate research attributes.  But once written, the Times added the cherry of journalistic malfeasance by completely botching the simple arithmetic of averaging two percentages.

One can only conclude on this sample size of one, that the New York Times is no longer a journalistic enterprise with professional standards.  They are profoundly unserious, innumerate and misleading.