The bane of inaccurate headlines

Before you get outraged and leave a comment, read the article.

YOU may have seen a headline sometime last week on a news website that said, “Childhood Vaccination Will Not Be Made Compulsory: Malaysia DPM Wan Azizah”. Given that my article last week was about the dangers of not vaccinating, it’s obvious where I stand on this issue.

Many other Malaysians online agreed with me as well, and asked how Dr Wan Azizah could be both a doctor and mother, and yet give leeway to those against vaccination. At least she isn’t the Minister of Health, they said sarcastically.

The thing is, everybody who criticised her kind of got it wrong. Not about the vaccination part, but about her not wanting to make it compulsory.

The headline did not match the quote attributed to her in the article, “We encourage parents to have their children vaccinated, but right now we cannot force them as it is their choice”.

And then, just in case you weren’t clear about her intention, she added: “We would like for parents to protect their children by vaccinating them before they catch the disease”.

There is a huge mismatch between the headline and article, and with it comes the outraged-yet-misplaced shouting, usually online.

How bad is the problem? An article from Forbes from last year was headlined “59 Percent Of You Will Share This Article Without Even Reading It”. It seems that people have low attention spans, exacerbated by the fact that just sharing takes less effort than reading articles in depth.

This is obviously bad. The article goes on to say that it contributes to the echo chamber effect, where people repeat what they’ve heard just because they like how it sounds, and don’t dig deeper.

This is partially made worse by the fact that the headline for this article doesn’t actually reflect the content. The headline says “59 percent of you will share this article without reading it”, but the content says “59 percent of all links shared on social networks aren’t actually clicked on at all”.

They’re not exactly the same thing. I frequently read an article, I share the link on Twitter, but nobody clicks on it. Living proof that having an unclicked article just means you’re unpopular, not that nobody read it. There’s also the situation when the same link is shared by multiple sources, and you only click on one.

In fact, the article takes these conclusions from a research paper, and the relevant sentence there is “We estimate that a majority (59%) of the URLs mentioned on Twitter are not clicked at all” – and then you realise the article confused “Twitter” and all “social networks”. Mark Zuckerburg would like a word with them, I suspect.

I do appreciate that research papers are not the easiest things to read, and the challenge is always how to summarise it succinctly but accurately. And for me, preferably without being over-sensational.

Unlike this one: “Race More Important Than Skills When It Comes To Jobs, Study Finds”, read one headline recently. It may not surprise you that this headline also doesn’t accurately reflect the contents.

It was referring to a study by think-tank Cent-GPS identified more than 500 entry-level jobs, and to each sent four CVs of men of different races, and three CVs representing three women. Their results indicated that a Chinese candidate was more likely to be called for an interview compared to the woman.

So, a more accurate headline would be “First-time Chinese jobseekers more likely to be called for interview than other races”, but I do admit it lacks the zip and sting of the original. At least the content seems correct from my understanding, and probably is directly taken from the original report.

But, guess what? The study has issues. I think there’s no issue with the conclusion that the CVs from the Chinese candidates attracted more calls for interviews. But the CVs were not exactly the same. For a start, they looked like they were of different colours. It seems colour choice matters; did you know that wearing red helps you win in sports?

I mean, it’s not too difficult a fix: If you really must have them in, say, two different colours, then make two versions for each candidate, and then at random which to send to each employer. Then it’s easy to check if race makes a difference, regardless of colour of the CV.

Perhaps more seriously, another difference was that the CVs did not list the same universities, and that clearly could impact selection.

The thing is, I myself cannot provide sources for these differences in the CV. The study by Cent-GPS isn’t incredibly forthcoming about the details of their study, and I only found out about some of the details because somebody put a photo up on Twitter showing CV examples (the tweet has since been deleted). I even sent a tweet to them asking for details, but I suspect it went unclicked.

Within a few days, several articles critiquing the study appeared, including one by Lee Hwok Aun and Dr Muhammed Abdul Khalid who had themselves written a seminal paper about racial discrimination in Malaysian university admissions. Somebody online commented that it was “probably the world’s gentlest peer review”.

Fundamentally, the conclusion was that although we recognise there are race-related issues for jobseekers, the methodology of the study has raised many questions, and shouldn’t be taken at face value – and probably not at the value of the headline.

I think we were lucky that the press has decided to go in-depth into this study, and is quick to recognise that scepticism is healthy.

That isn’t always the case. Sometimes we let politicians discuss issues through politics and the whole thing becomes quite shouty.

Informed debate – bouncing ideas back and forth – is actually for the better, hopefully in the end resulting in positive change.

But the thing is, it’s tough to even start this process if all you read are the headlines.