What's wrong with this article is that it's completely misleading. Let me analyse....

Hopes that the UK's IT jobs market might be making a recovery were dashed today following confirmation that the number of vacancies suffered their worst drop since the economic slowdown began 18 months ago.

How can vacancies drop? There are two possible answers.

1. People are losing their jobs and there are no new ones for them to fill.
2. People are not losing their jobs but new graduates are finding that there are few or no vacancies available.

The slowdown started 18 months ago but during what time period did the "worst drop" occur? Last week? last month? Was it getting progressively worse? Or was it actually improving and then just suddenly slumped. The last would actually indicate a rise in jobs in the IT section.

Example. Slowdown starts. IT jobs vacancies are at 7% above demand. Nobody to fill them cos they all have jobs. New recruits sought from abroad and from 3rd level. Most jobs filled. Vacancies drop to 3%. This is a "drop" in vacancies but it can't be construed as a negative one because jobs are being filled. The fact that there are vacancies at all implies that the job market is still healthy.


The number of vacancies slumped 38 per cent in the last three months of 2002 according to the CWJobs UK Quarterly Regional IT Skills Index.

This is bullshit. The word "slumped" implies that there is a crisis. Not so. It really means that the market has peaked and that there are fewer jobs than there were at the start of the rising economy. And 38% in the last three months? 38% of what? Does it mean that there are 38% less ads in the paper? Or 38% less jobs available? Or does it mean that 38% of jobs have been lost? This can beconstrued two ways although the figure above is used in the negative sense:

1. 38% less vacancies. Jobs are few.
2. 38% more jobs filled and unemployment is lower.

The first case could imply that the market demands a consistently high level of vacancies to encourage confidence and to keep the market healthy. The second case implies that the employment level is higher and that the available vacancies are becoming more specialised and localised. Not necessarily a bad thing.

It seems the dip has put an end to hopes for an early recovery in the IT jobs markets, especially after earlier figures suggested that the rate of decline was slowing down. It sounds like people began filling the available positions.
The East Midlands was the worst hit region in Q4 2002, with the number of vacancies dropping by 45 per cent, closely followed by the West and Wales (43 per cent).

Bullshit again. They are using the figures for less vacancies to imply that unemployment is increasing. These figures are useless if not used in that context. If vacancies are dropping by 43% but people are still getting jobs without becoming unemployed then what's the problem?

At least he admits that people are filling the available positions.

Overall, all regions suffered a fall in available jobs with every part of the UK seeing over a third less IT jobs advertised than in Q3 2002.

This is what's called a glut. IT should be discouraged as a subject and the supply of other in-demand skills should be taught in Universities and colleges. It's the way it's always been done.

There are jobs there but not if new graduates are continually churned out.

For the whole of 2002, the IT jobs market fell by 76 per cent. The Midlands suffered an 80 per cent fall in IT job vacancies while Central London racked up a 62 per cent decline in advertised IT vacancies.

This is playing with numbers. IT jobs market falls by 76 per cent. Nationally? But 76% of what? The figure last year? The usual amount of job vacancies you find in the Sunday newspapers? It's an empty figure and makes no sense without a comparison.

It's an interesting observation that most people take statistics at face value without actually questioning what they really mean or imply. If things were really that bad in IT in Europe then why don't we hear of mass unemployment? Because it is all bullshit. You can play with figures and percentages and get them to mean anything you want. For example, if you have five large IT companies who employ 10,000 each and then you have 45 unstable dot coms which employ five geeks each and those 45 fail within the year then you could statistically say that 90% of computer companies failed during the past year and you would be inclined to believe that the IT business was in dire straits whereas only 225 damned fools would be out of work and 50,000 people's jobs were secure. The statistic doesn't do much for your confidence but the real story speaks louder.

I'd like to know in which areas the statistical falls are taking place. Are small companies failing quicker whilst multi-nationals are recuriting? And are they failing quicker than those larger companies can recruit? That would be one way to explain the fall. There are probably more than a thousand other explanations. One thing is for sure - the IT market is still strong enough to employ.

When I was doing my degree in the 90s one of my subjects was statistics and the first thing we were taught was how to gather and display information correctly. If it isn't done so then it can be misleading. You can also lie using statistics and twist the information to show the opposite of what it was intended to depict. Not difficult when you look at the above article which I completely disagree with.

My bible for that class was the following which is a short and readable book I'd recommend to anyone who doesn't want to get fooled by opinion polls again.

AUTHOR :Huff, Darrell
Calman, Mel
TITLE :How to lie with statistics / [by] Darrell Huff ; with pictures
by Mel Calman.
IMPRINT :Harmondsworth : Penguin, 1973.
COLLATION :124p. : ill., maps ; 19cm.
SUBJECT :Statistics
SERIES-Title :Pelican books
GENERAL NOTE :Originally published, London: Gollancz, 1954.
ISBN / ISSN :0140213007 (Pbk) :