Box Turtle Bulletin

Box Turtle BulletinNews, analysis and fact-checking of anti-gay rhetoric
“Now you must raise your children up in a world where that union of man and box turtle is on the same legal footing as man and wife…”
This article can be found at:
Latest Posts

Beware: Journalists Reporting on Science

Rob Tisinai

September 23rd, 2010

You’ll be hearing our opponents trumpet a new article from the UK Guardian. The headline reads:

1.5% of Britons say they are gay or bisexual, ONS survey finds

The article messes things up pretty fast. Here’s the second paragraph:

The findings, based on interviews with more than 450,000 people – the biggest pool of social data after the census – show that an estimated 481,000 people regard themselves as gay and a further 245,000 – mainly women – say they are bisexual.

And here’s the third:

The estimate puts Britain’s gay community at 1.5% of the total population – much lower than the most commonly used estimate of 5% to 7%, which was cited by ministers introducing civil partnership legislation and implied a non-heterosexual population of 3.5 million.

See the problem? The article assumes the percentage who say they are LGBT equals the percentage who are LGBT. Is that reasonable? No. One obvious objection is that some people (especially the elderly) aren’t willing to tell a stranger they’re gay. But the article is even sloppier than that. Here’s a snapshot of the data:

Category Percentage
Straight 94.8
Gay/Lesbian/Bi 1.5
Other 0.5
Don’t know/refused to answer 2.8
No response 0.5

You see that? The study failed to identify the orientation of 3.8% of the population. There’s only one way to conclude, “The estimate puts Britain’s gay community at 1.5% of the total population,” and that’s to assume none of these people are gay, lesbian, or bi.

Is that reasonable? Given the way straight people flaunt their sexual orientation, and given the continued existence of the closet, it makes sense to wonder if a good chunk of that 3.8% is gay, lesbian, or bi (not to mention transgender).

Unfortunately, based on my experience with the right wing echo chamber, our opponents will repeat the 1.5% figure over and over until they’ve forgotten where it comes from. But that makes sense. Sloppy numbers are the best defense of sloppy thinking.



September 23rd, 2010 | LINK

“not to mention transgender”… hmmm i wonder why you write this.

though myself cisgender, my understanding is that being transgender is not on the same spectrum as being, say, straight or bi (if, of, course, such spectrum does exist; if you don’t like the term ‘spectrum’ , use ‘category’ instead if you must).

case in point: transgender individuals can be straight, bi, or gay. it’s not like they’re transgender and that’s their sexual orientation….
someone cares to educate me on the matter?

Rob Tisinai
September 23rd, 2010 | LINK

David, I wrote that for exactly the reasons you mention. I know transgender people who variously say they are homosexual, heterosexual, bisexual, or a sexuality that they will not label with ANY of those categories. The survey doesn’t seem to handle that adequately, so the transgender population is a mystery as far as this survey is concerned.

September 23rd, 2010 | LINK


Emily K
September 23rd, 2010 | LINK

I think that kinda makes sense.. I mean, if an intersex person refuses to identify with any single gender, what does that make their sexual orientation when they couple with someone of exclusively male or exclusively female persuasion? Or another intersex person?

Timothy Kincaid
September 23rd, 2010 | LINK

The questioning involved showing people a card of options and asking them to indicate which category they fitted into. As a result, the ONS is highly confident about the results.

Oh My Dear Sweet Baby Buddha!! Is there a less accurate way to get an honest response?

This is WORSE than tick-the-box. They actually are asking people to say out loud to a total stranger that “I’m in the ‘gay’ category.”

Timothy Kincaid
September 23rd, 2010 | LINK

And lets look at the age breakdown:

16 – 24 1.9%
25 – 44 2.0%
45 – 64 1.2%
over 65 0.6%

Anyone see the obvious flaw in this “research”?

September 23rd, 2010 | LINK

As much as I hate to perpetuate a stereotype, given the population I have to wonder not only how many “people (especially the elderly) aren’t willing to tell a stranger they’re gay”, but how many British people are willing to discuss sex with a stranger.

As silly as it sounds I’ve known quite a few British people who’d respond to a survey like this with, “A little less sex, please. We’re British.”

Timothy Kincaid
September 23rd, 2010 | LINK

Not to mention the obvious socio-economic pressures of self-identification to somebody holding a clipboard. Fully half (49.1%) of those who identified as gay were in “safe” managerial positions (as opposed to the average of 30%).

This is a HORRIBLY conducted survey.

Brian Pengelly
September 23rd, 2010 | LINK

I guess a better question is what does it matter? Are gay people less deserving of protection and their civil rights if they make up 1.5% of the population than if they make up 7% or 10& or 50%?

Even if we make up 0.005% does that make it okay that we are treated unfairly? Are Jews less deserving or respect and protections because they are outnumbered by Christians? Are Native Americans less worthy of protection because they are outnumbered by Hispanics? How many people does a minority have to have in it before it is valued?

September 23rd, 2010 | LINK

And how many of those claiming to be straight were lying?

September 23rd, 2010 | LINK

Self-reporting is the weakest form of polling. It is the most prone to lying, exaggeration and omission.

Self-reporting is the reason that teabaggers, for instance, can come out of NYT/CBS polls as considerably MORE educated (including advanced degrees), more financially savvy and with higher incomes than the average American, despite every appearance to the contrary.

Every poll is self-reported, to an extent. Polls that ask about highly emotional issues but don’t probe the answers, should be taken skeptically.

Chris McCoy
September 23rd, 2010 | LINK

The ONS Survey can be found here

Their analysis of the “Sexual Identity” question breakdown is here.

Self-reporting Surveys will always be under-reported. The official analysis doesn’t appear to even consider under-reporting.

Mark F.
September 23rd, 2010 | LINK


The liberals are obviously turning our young people gay! ;-)

September 23rd, 2010 | LINK

I think the only way to find out the real number of LGB people is to force EVERYONE to participate in those experiments where they test arousal levels.

September 23rd, 2010 | LINK

I have come to the impression that with the exception of a few ethnic groups and regions that have higher rates, rates of homosexuality &c are pretty uniform worldwide.

This survey is awful, though. When you’re trying to poll for identifications that individually run around 1-2% in the population and aggregate to 5% (8% tops), allowing 3% “don’t know”s defeats the whole endeavor.

Paul Mc
September 24th, 2010 | LINK

I’m afraid the headline 1.5% gay is likely to stick. It is consistent with other population surveys in the UK and France in last 10 years.

A journalist not looking for cheap filler would then ask why, if there are only 750,000 gay or bisexual people then how come ~1.5Million people (wish there was some collective data about this – just my guess from news reports) attend Pride events in the UK? With a demographic of ~18-50 based on my observation.

If only two thirds of those are in fact LBGT and also that this demographic age range is 2/3 of the whole LBGT community (including those not yet out/adult) and finally that 2/3 of this demographic actually bother to attend any pride, that puts a UK LBGT commuity (those largely identifying as LBGT openly) at 2.25million or 3.75%.

Does anyone have a better way to do this than my 2/3 rule?

September 24th, 2010 | LINK

Please note how they conducted the survey: “In both cases, the sexual identity question was administered to all members of the household aged 16 or over who were available at the time of interview to provide their own responses”.

Does anyone really believe that they could survey a household and obtain valid responses?

September 24th, 2010 | LINK

The problem is with the ONS survey, not the Guardian’s reporting of it.

Priya Lynn
September 24th, 2010 | LINK

Clay, the problem is with both.

September 24th, 2010 | LINK

“Gay networking websites Gaydar and GaydarGirls say that estimates of 1.5 per cent of the population being gay or bisexual cannot be true as they have 2.2 million members in the UK.”

Helen in Ireland
September 26th, 2010 | LINK

“The findings, based on interviews with more than 450,000 people – the biggest pool of social data after the census – show that an estimated 481,000 people regard themselves as gay and a further 245,000 – mainly women – say they are bisexual.”

My problem is with the basic maths – how can a survey find that 31 000 people MORE than they surveyed say that they are gay? Nevermind the 245 000 who are bisexual ?

Or was there a typo in reporting the true figure…?

Leave A Comment

All comments reflect the opinions of commenters only. They are not necessarily those of anyone associated with Box Turtle Bulletin. Comments are subject to our Comments Policy.

(Required, never shared)

PLEASE NOTE: All comments are subject to our Comments Policy.