Computers: Young Essex women conform to type: William Hartston enumerates some of the revealing facts that can be extracted from the CD-rom of the 1991 Census

Where were you on or about the evening of 21 April 1991? If you are female, in your twenties and work as a secretary or clerk, chances are you were at home in Essex. Or you may have been visiting friends in Oxfordshire. But if you found yourself without a bath or shower you were more likely to be in Cornwall.

The 1991 Census is now available on CD-Rom - a snip at pounds 2,500, but only if you are a state- funded higher education institution or public library in the UK - and provides the richest possible seam of information for the serious researcher or the demographically prurient. With England and Wales divided into counties and Scotland into regions, then subdivided into districts, which divide again into wards (postcode sectors for Scotland) then further into enumeration districts (output areas in Scotland) data are available for almost 150,000 areas.

And with between 9,000 and 20,000 items for each area - fewer items are available for the smallest areas - the whole forest of information has spread to take up eight CDs. Thanks to sensible allocation, however, there is little disk-shuffling needed in its operation for most natural purposes. Of course, if you want to compare the number of single-parent families in Strathclyde with that of Southend-on-Sea, you will need to start the whole system on disk one, then load disk seven for the Scottish data and disc eight for Beds, Bucks, Berks, Essex and other far- flung points of the commuter belt.

Which bring us, somewhat tenuously, to 'Essex girl'. Of the 1,468,656 Essex residents present on the night in question - of whom, incidentally, 0.032 per cent were born in China - 51.35 per cent are female and 39.8 per cent of the females aged between 25 and 29 are single. That is not a remarkable figure. In Inner London 69 per cent of women in their late twenties are unmarried.

What makes 'Essex girl' stand out is her occupation. Table 74: Occupation (10 per cent sample) lists the jobs of 7,427 Essex women aged 20-29, under 11 headings. While there are 924 'Managers and Administrators', 676 in 'Sales', and only 165 in 'Craft and Related', these figures pale when compared with the 3,097 'Clerical and Secretarial'.

Forty-two per cent of economically active Essex women between 20 and 29 are secretaries, the highest rate in the land. For Cumbria, the comparable figure is 25 per cent and in the Isle of Wight it drops to 24 per cent. The most secretary-rich environment of all is Castle Point, Essex - Canvey Island and Benfleet - with a startling score of 52 per cent.

Few people visited Essex on the night of 21 April 1991. Only 31,554 people in that county on the night of 21 April 1991 were non-residents. Only Derbyshire attracted a lower rate of visitors. Oxfordshire was the most popular place to spend the night in someone else's establishment, with 7.13 per cent of those surveyed having come from somewhere else, while Derbyshire and Essex were down at 2.09 per cent and 2.15 per cent respectively.

The most atypical areas in England and Wales seem to be Inner London and the Isle of Wight. Inner London has the highest rate of babies who have not yet reached their first birthday and by far the highest rate of people aged between 20 and 24, yet the lowest rate of 16 to 19-year-olds. It also has the highest percentage of people born in China (0.146 per cent), the highest rate of unmarried people, both male and female, and the highest proportion of households without a car. It is also the only county-sized area with more unmarried females than males in the 25-29 age group. (Suffolk has the highest rate of unmarried young males.)

The Isle of Wight, however, has fewest babies under one, fewer people under-16 than anywhere other than Dorset, the highest proportion of women outside Merseyside, the most over-80s outside East Sussex, the most households without a bath or shower outside Corwall and is second only to Oxford in attracting visitors. It is also the only place in the country where none of the 16 or 17-year- old males are married.

Most of the above details could only be arrived at after a little manipulation of the raw figures, but once the data has been transferred to an on-screen table, simple mathematical operations are easily performed thanks to an 'Add Expression' command, which permits additions to a table to be made with user-defined items formed from existing units.

The results may then be displayed and printed in a variety of forms including bar charts, pie- charts, histograms and 'choropleth maps', where different shadings or colours indicate the distribution of the data figures.

The instructions, like so many software manuals, are easy to follow if you know what you are doing, but are difficult to understand from a standing start. As usual, familiarity comes with playing with the system, but I would have saved a great deal of time if I had started by reading Appendix D: Exercises, which takes you methodically through examples of setting up tables, finding data and manipulating it to obtain the desired results. The opening chapter on 'Installation' does, indeed, end with a suggestion that you 'refer to the following chapter to find out how to start the 1991 Census on CD-rom and/or refer to the exercises in Appendix D', but for most users the sign to Appendix D would have been better put in large capitals.

The tempting game is to try to identify one's own enumeration zone, but the compilers have been sensitive to the matter of Census data secrecy. All figures are presented as aggregates, so it is impossible to gather information about a single individual. You may discover that there are four women in Brentwood aged 20-29 working in 'Craft and Related Occupations', but you cannot find out their marital status or whether they have their own baths or showers.

One may ask for progressively more information about Inner London, Islington and Highbury, but beyond that level - where there are 8,290 residents of whom only 2 of the 17 over-90s are male - the sections are identified only by code. Even if you decide that your district must be ED01AFFG21 on the grounds that it is the only one with both a married female aged between 16 and 17 and 6 males born in the Irish Republic, you could still be wrong. For the Small Area Statistics (SAS) and Local Base Statistics (LBS) are not quite what they appear:

'Anonymity was further protected by the modification of . . . data for small areas. For the SAS, this involved adding +1, 0 or -1 in a quasi-random pattern to each cell in the 100 per cent data . . . For the LBS, the same procedure was used twice so that each cell was altered by the addition of +2, 0 or -2.'

So the sole Plant and Machine Operative, Female, aged 20-29, in Brentwood, may not exist at all. Or she may be another quasi-randomly misallocated secretary.

THE 1991 Census on CD-rom includes 8 CD-rom disks with installation software on two 1.4-megabyte floppy disks and a 200-page user manual. It is available only to state-funded institutions and public libraries at a price of pounds 2,500. An expanded Health Service Edition is priced at pounds 3,995. The Census is produced under licence from HMSO by Chadwyck-Healey, The Quorum, Barnwell Road, Cambridge CB5 8SW; tel 0223 215512.

The choropleth map used above charts percentages of married teenagers throughout England and Wales, ranging from 0.71 per cent in Merseyside to 2.25 per cent in inner London. The lightest shading represents the highest rates.

1991 LBS/SAS statistics Crown copyright 1991 All rights reserved.

Digital boundaries for England and Wales OPCS, OS and GDC 1991.