The Kiwi accent is apparently the world's sexiest ... but is it really? Stats NZ talks dirty on the validity of that poll

The Kiwi accents has been named the world’s sexiest in a poll, beating the likes of the Irish and French.

COMMENT: You probably read the headlines this month confirming something Kiwis have long known: The New Zealand accent is the world's sexiest.

Most media outlets, including here at 1 NEWS , reported the results of a survey conducted by Big 7 Media - a well-followed travel site - whose readers had apparently voted "New Zillund" into the top spot.

What set this listicle (list-article) apart from the usual crop of travel content was that this wasn't just the opinion of the writer, apparently.

These results were derived "from a sample survey results of our 1.5million social audience".

Read more:  Famous film scenes recreated with ‘sexy’ New Zealand accent

So how was it that a global poll produced this result, I wondered. How exactly did they get that number?

All good polls should really explain how they got their information - referred to in statistical circles as their methodology - so I asked the company that did it.

Some media outlets use polling companies with trained statisticians to conduct their research, but in this case, the company that conducted the polling for Big 7 Media was Big 7 Media.

Sarah Clayton-Lea, Big 7's head of content, was happy to answer my questions, telling me "it's great that the accent results are sparking such interest in Big 7 Travel.

"Yes, we can confirm the validity of the World's Sexiest Accents poll ... the survey was conducted in December 2018 as part of a series of audience surveys on Instagrammable destinations, accents and more," she said.

"We don't run organic polls on Facebook or Twitter, as we have found these to be skewed towards an American audience - our surveys are promoted on several of our channels, mainly Instagram stories.

"We created a questionnaire with a list of 50 recognised accents and asked our audience to vote for their favourites - this questionnaire was pushed to our readers via our email database of 12,000, Instagram Swipe Ups on our three Instagram accounts, and an on-street survey in London and Dubai of 1500 people."

Read more:  Invercargill librarians have fun with NZ's 'sexiest accent'

The total number of people who responded, Ms Clayton-Lea said, was 7300. So let's break this down.

The poll was made up of answers collected from 5800 of Big 7's Instagram followers and email subscribers and 1500 randoms on the street in London and Dubai.

Among those, a total of 21.7 per cent chose Kiwi as the sexiest accent - that's a grand total of 1584 people who deemed our accent the pick of the bunch.

I took this information straight to New Zealand's sexiest arbiter of statistical information - Statistics NZ.

As the Kiwi accent is apparently the sexiest in the world, these librarians turn on their phrases.

STATS NZ TALKS DIRTY

"In general people should ask how surveys were taken, how many people were polled, and ask for the margin of error - if that's big, it is likely to mean the results are potentially unreliable, if not misleading," a Stats NZ spokesperson told me, sexily.

"That's the statistical equivalent of checking if someone already has a partner before asking them out on a date.

"For a survey to be considered statistically sexy -  that is, robust and reliable - it should have a good sample size. The relationship between sample size and population size isn't linear – so a small sample in a whopping population could still be pretty robust."

OK - so it's not always about size - got it.

"A survey of, say, 10000 people in New Zealand could be a ruggedly handsome sample size, and would have a relatively small margin of error - a sample of 7300 in a population of 1.5 billion may be the dating equivalent of a weedy whimp ."

So, considering there are more than a billion English speakers in the world, I took this to mean that this survey was somewhat of a statistical 'left swipe' - but it probably has a good personality.

"It would be good to know the demographic characteristics of this sample to determine if some of these characteristics skew the results of the responses - it should ideally be a random sample - people chosen by chance - not a sample of people that could skew the results.

"For example, if you were asking people about their use of Tinder, you may get quite a different result depending on whether you went into a bar in central Wellington, or a retirement village on the Kapiti Coast."

The spokesperson then began to seduce me using terms like "coverage bias" and "adjustment on inferring to the target population", and before I could catch my breath, they whisked me off me feet by telling me:

"It may be as scientific as asking a bunch of friends how sexy they find your voice."

But the New Zealand accent remains the world's sexiest, doesn't it?

"That statement is, without question, true," Stats NZ said with a wink, "based on the voices of our national heroes - from Fred Dagg and Billy T James, to Sam Neill and Cliff Curtis.

"But unless the points raised above were addressed, then the generalisation only reflects the views of those who chose to respond."

News tip or more information? Email Luke Appleby or 

SHARE ME

More Stories