Breakfast News reaches 30. Big event. Big data. Big breakfast.

I am here today, brothers and sisters, to celebrate the 30th TARGETjobs Breakfast News which was held last week in London in front of an invited audience of recruiters, agencies and universities. This event, which is sponsored by Work Group and held in association with AGR, has become a well-established forum for sharing ideas amongst senior members of the graduate recruitment community. It is truly a hot ticket with a hot breakfast. If you get there early enough. Otherwise it’s a bruised pastry.

But before I address the serious content of the morning, I want to make an announcement and a confession.

I am a fully paid up member of the world’s largest religious community: The holy order of lapsed Catholics. I mean, why would anyone want to be a practising Catholic anyway? All that guilt, incense and management infallibility is too much to live with – until, that is, the first notes of the last trumpet sound and we hasten back to the one true faith, tails between legs, looking for the forgiveness that redeems a lifetime of debauchery, lies and greed.

Having sat in the audience at TARGETjobs Breakfast News listening to the presentations, I felt moved to confess (it’s another thing that Catholics do). I therefore confess the following:

1. In my career, I have been guilty of misusing research findings to support a particular viewpoint, even when that viewpoint has not necessarily flowed logically from the findings of the research.

2. Although I do know the difference between ‘research’ and research, there have been times when I have wilfully blurred the distinction in front of an audience.

3. I have said the words – ‘Research has shown that…’ – when it hasn’t. More than once.

4. I have called something Research when I should have said Survey.

The theme of the 30th TARGETjobs Breakfast News was Big Data and each presenter treated the subject in a different and interesting way. Our resident economist, Dennis Turner, looked at the difficulties of economic forecasting in an era when there has never been more data available. I found this presentation rather reassuring. Stephen Isherwood, AGR CEO, focused how internal and external data can help recruiters and the eminent commentator Lord Daniel Finkelstein put big data in an historical context.

But it was Marcus Body, Head of Research at Work Group, who made come over all Catholic.

He’s the Head of Research for God’s sake and here he was biting the hand that feeds him. A lot of research, I paraphrase, is not worth the paper it’s written on. As always in modern life, it pays to exercise a certain degree of scepticism when people are trying to sell you something and a shedload of cynicism when their words are based on research.

It was a detailed presentation but I have extracted some important questions which you may like to ponder:

1. Who did the research or who promoted the research?
This is the most important question to ask in order to understand the outcome and recommendations. Does the researcher have a commercial or political reason for commissioning or promoting the research? A Daily Mail poll which shows that 97% of the population hate asylum-seeking foreigners can only be evaluated after understanding the culture of the messenger. Not that I’m making any judgement here.

2. How big is the sample?
A surprising amount of research findings are based on a laughably small sample. Samples need to be large and diverse enough to be statistically valid. A recent research finding on a cardboard cup told me that Costa Coffee was the nation’s favourite coffee. It may well be but, in small print on the other side, it seemed to be based on a poll of fewer than a hundred people. Similarly, a headline-grabbing finding that 75% of people are angry about something is undermined when the total sample size is 12. You’d get fewer headlines if you said: 9 of the 12 people who replied are feeling angry.

3. Is it the right sample?
Generally speaking, in your professional capacity, you’re more interested in research that will give you information and insights to help you become more successful or more knowledgeable. So it pays to understand who exactly is in the sample and to judge how relevant they and their views are. It’s not unusual for the presenters of research to gloss over the demography of the sample.

4. Does the data support the conclusions?
It pains me to say this, but some research is done to support a conclusion that has already been made. Pay particular attention, therefore, to the correlation, or lack of it, between the data findings and the conclusions/recommendations. They don’t always sing from the same hymn sheet.

5. Was all the data used?
It’s an old trick to select only the data findings that support the general thrust of the argument. If certain data is not analysed or used, it’s nearly always a bad sign and begs the question: why didn’t you mention it then?

This is a very outline summary of Marcus’s presentation and you won’t be surprised to learn there are even more ways to tell the difference between research and ‘research’.

I want to stress that not all research is questionable. I guess the point is that not all research is research at all. It’s marketing, which is fair enough, but marketing that’s given spurious importance and glamour by masquerading as research. Or it’s a simple survey re-badged as research. It’s not, in truth, too difficult to see through.

In my experience, the only way to identify thoroughly researched data is to ask one simple question:

Are people (governments/corporations/special interest groups) sitting on research or delaying its publication?

If the answer is yes, you can be 100% certain that it’s kosher.

You can have a look at all the presentations from this and previous events here:

http://gtimedia.co.uk/expertise/targetjobs-breakfast-news

May God forgive me.