My Life Story 3- Uncharacteristic Decisiveness

Not relevant at all, just thought I'd stick it in to brighten things up. It's my most recent birthday cake; 24 happens to be my favourite number as well as, this year, my age. Not that I'm anything other than a person with strictly rational and scientific beliefs, of course...

As I said, I returned to the UK without a good idea of what I was going to do. I went back to live with my parents while I thought, in a small village in Gloucestershire*. It was around this point that I realised that my future career was not going to come to me in a flash of inspiration, filling me with enthusiasm and determination for my new path. No, I was going to have to research options, find out about stuff, and have a good long think about what characteristics I wanted my working life to have.

The results of these deliberations were chiefly 1) having missed maths while teaching English I wanted to do something in that area and 2) I liked the idea of working for the government- I wanted to be part of something usefull and important- not necessarily a big part of something important, in fact preferably a small part, but anyway, a part. There were other things on the list such as 'must have proper lunchbreak where you can take time off to go and buy a sandwich and sit in a park' and 'must involve large amounts of proper work rather than having to pretend to work through having nothing to do', but I could tell they weren't going to be so easy to find out about before starting the job, and in any case were probably a bit much to hope for, so I confined my attention to the first two points. Also to bear in mind was my list of three not-if-I-can-possibly-help-it professions: accountancy, actuarial work, and teaching.

I found three jobs using maths in the government: being a mathematician at GCHQ, working for the Audit Commission (perilously close to accountancy, but redeemed slightly by being about making sure public money was being spent properly which is obviously a worthy aim), and becoming a statistician with the Government Statistical Service- which, when I found out about it, immediately appealed. I liked the idea of using mathematical principles to collect and analyse data which could then be used as the basis for making important decisions (so long as I didn't have to make those important decisions myself)- it sounded very interesting, since I've always liked finding out data about pretty much anything except sport, and transport, healthcare, education and so on are areas I've been wanting to know more about; and usefull because without access good figures there's no way the government could be expected to make important decisions well. The other two I was applying for mainly as back-up options since one application only seemed rather unwise: I wasn't so keen on them- the Audit Commission didn't sound that interesting, and the GCHQ job sounded too maths based- I wanted a job that also involved dealing with the real world. From what I could gather, the GCHQ job was about researching ways to break codes, which would certainly be interesting, but I wouldn't want to be doing only that. The other problem was that I was pretty sure I wouldn't be able to do it- by the end of my four year maths degree I realised that while I was good at learning and understanding concepts already discovered by others (up to a certain level), I had no mathematical research talent whatsoever. Which wasn't a problem as I felt that it would be too narrow and not real world enough to really interest me anyway- but it did mean that I didn't feel it was that likely that I'd get anywhere with the GCHQ application.

As it turned out, the Audit Commission rejected me first, at the receipt of application stage, but I wasn't that bothered as if I was going to be unsuccessfull, I was glad that it wasn't yet with the Government Statistical Service. Then I was invited to a selection day at GCHQ and failed that- I'd have been worried for national security if I hadn't because I did very badly on the test, not even managing one question. Again, I wasn't really upset because I still had the possibility of the GSS. I got through the online tests, and the 'e-tray assessment centre' (a repeat of the online verbal reasoning and numeracy tests plus multiple choice decision making based on a series of fictional emails), and even the Fast Stream assessment day which, like the previous parts, was not just for statisticians but for Fast Stream Civil Service, Economics and Diplomatic Service candidates. It was designed to test a variety of skills such as communication, decision making and working with others, and it was based around fictional scenarios that seemed more appropriate to the main Civil Service. I started to feel a bit unsure because I only wanted to be involved with information collection and analysis, not decision making and project development which was all very interesting but really not my strong suit- I talked to some people currently on the scheme at a careers fair back in the autumn and they had said that you couldn't just stay in statistics but would inevitably be promoted away from it- though they suggested that no-one would want to stay in statistics anyway. Looking back, I wondered if I should have paid more attention to this. But I couldn't think of anything else I wanted to do so pushed these doubts down and continued to work for the final, specifically statistics assessment- I'd been studying statistics from textbooks because, apart from a tiny bit at A-level that was compulsory even if, like me, you weren't taking the statistics courses, and another small bit in the first weeks of my degree which again was designed to give even those trying to avoid it a little bit of knowledge**, really the last statistics I'd done was at GCSE. I managed to get a fair way through the textbook, but not far enough, and I (just) failed the final selection because, even though the assessors were really nice and made allowance for my lack of knowledge, trying to guage aptitude instead, I just didn't know enough for them to judge my potential or otherwise (as they explained).

This time I was quite upset about being rejected, because this was the job I had wanted so much. I had a few weeks where I was totally unsure of my plans again, since initially I decided I should think again- but when considering other jobs, my immediate gut reaction was always that they were rather lacking in statistics, so I then realised that I still wanted to work in the area- and I would have to accept that I couldn't hope to be accepted as I was but would need to do a masters in statistics. Going back to university had always been attractive, but I had felt I should get a job and earn some money. Now I came to realise that it really was the option that made the most sense. My parents were prepared to loan me my tuition fees and living expenses for the year, so I decided to go for it and immediately felt very happy with this plan.

I researched courses at many universities online; given all the money in the world and all courses being equal I'd have gone back to UCL without even thinking, but in the interests of minimising costs I was hoping to find somewhere outside London, whilst keeping to the main priority of finding a good syllabus suited to my needs. When looking into places to do my first degree, I can remember reading subject lists filled with things like 'Analysis', 'Rings, groups and fields', 'Further Linear Algebra' and so on which didn't mean a thing to me; I asked my dad and my teachers about what some of them meant but I have to say I didn't really grasp any of it enough at the time to make even an attempt at basing my decision on the course content. This time, I had a better idea- I'd come across most of the topics during my study that year, not from having gone into them in detail in working through my textbook, but from looking them up in a dictionary of statistical terms that I bought quite early on and which was great for brief but understandable basic explainations of things. As well as looking for Time Series*** and other areas mentioned as usefull on the GSS application form, I was hoping for something about survey techniques because I realised in the written part of the final test that I didn't know anything about this beyond common sense. Of all the universities I looked at, the only one that had a course in this was LSE. It also had courses in Time Series and pretty much everything else I was looking for. So that was how it became my first choice. It's an expensive place, which isn't ideal, but I thought it would be a false economy to go somewhere cheaper with a less suitable syllabus, and I'm lucky to be able to borrow enough money from my parents to cover it, as well as getting £3,000 off the tuition fees (something called a Graduate Merit Award). And I can't deny that the prospect of being in London again was a happy one. Fortunately for me, even though this was April, there was no deadline for applications- it was open untill the course was full. I got mine in before this happened (though not by much) and heard some weeks later that I had been accepted.

This brings the reader pretty much up to date- it only remains to sketch out my plans for after this year. I'm going to apply to the GSS again, though I'll try and do some more research and find out how much stuff other than statistics would be involved. But I'm now leaning more strongly in the direction of other statistics jobs: maybe in the pharmaceuticals industry- they need statisticians to monitor all complaints developed by patients after drugs are licensed to pick up on the side effects that only occur in small enough percentages that they can't be spotted in the original trials, as well as, I imagine, for those original trials and other things- but I'm not quite sure about the ethics of this, because while this is obviously important and beneficial, I'm not sure of my position on animal testing which is generally a part of the process. So I'm also strongly considering market research firms and independent social research bodies whether they be firms or academic institutions- basically at this stage I'm interested in anything and I have to do a lot of research. Some of these jobs will probably be graduate recruitment schemes (the GSS certainly is) which will have deadlines in October or November for a job starting next summer, so I'll need to be ready to apply by then, in spite of being busy settling in and getting to grips with the courses, whereas others will be to come back to about a month before I would start (when I'll be in the middle of the rather scary sounding dissertation- timing, eh?)

*I lived there from the age of 1 1/2 till I went to university, and thought it was pretty isolated, the nearest shops being too far to walk to at a five minute car journey away, but after my experiences in Japan I readily admit that this is civilisation. After all, it's only 2 hours to London from here, whether you drive or go by car then train, compared to 5 hours not including connection times from my village in Fukushima to Tokyo, though it must have been about the same distance (it's those mountains, they get in the way)- and what I used to think was very inconvenient public transport, consisting of a bus running every two hours or so, making a half hour journey to a town with many shops, I now see as a great service compared to three buses a day to a place only distinguished from where I started out by having a train station, with the shops being two hours away. Yes, I know I went through all this in the previous post. I only mention it again because I forgot to say last time that, living in a rural area already I thought I knew what to expect in Japan and that I'd slot right into the way of life, whereas in fact it turned out to be completely different and to take some getting used to.

**I had, in fact, up to that year been avoiding it, or trying to- I hated statistics at school- because, though I didn't at the time realise that was the reason, all the examples in the textbooks were made up so you didn't have the interest of finding out something real, they were usually about really boring things like the distances thrown by a shot-putter or the heights of a class, and GCSE level statistics was so simple that even if it had been real data about something interesting, it wouldn't have been possible to find out much more than the average of various kinds and maybe a positive or negative corellation- fairly interesting, and enough for me to enjoy the couple of projects where we got to collect our own data (though I was so sure I hated statistics that I barely registered at the time that I was enjoying it), but nowhere near as fascinating as inference testing, working out how many cashiers are needed to serve a queue of people arriving at a certain number a minute, or any of the other things that form the most basic level of university statistics. Since I was so sure I hated the subject, I was very definite about not taking it at A-level (when for all I know now it might actually have become interesting), and UCL is one of the few universities with, apart from that couple of weeks at the beginning (which I still wasn't that sold on- but then it was mostly simple probabilities; I was probably finding it more interesting than GCSE (especially the conditional probability and the Poisson distribution), but it was having to fight against my conviction that I didn't like statistics and wasn't quite enough to manage it), no compulsory statistics courses- there were options in the third and fourth years that I didn't choose because of the aforementioned prejudice.

***Time Series: For example, suppose you're an ice-cream seller who has just put out some advertising and wants to see what effect it's had on sales. If it's now September, then comparing sales to last month you'd expect to see a drop as people normally buy less ice-cream in September than August, and so that drop doesn't tell you that your advertising is actually putting people off buying your product. In order to find out whether it has boosted sales and by how much, you need to compare previous Septembers. But it's actually quite a bit more complicated than that, because although that will tell you a fair bit, there are still general yearly trends- perhaps ice-cream sales are increasing generally from year to year and you need to know whether an increase compared to last September is because of the advertising or a general trend (or both). There are ways of adjusting the monthly figures by looking at previous years so as to remove the seasonal effect, which will let you compare August and September on an equal footing and reveal whether the advertising did produce an increase after all, but I'm not sure of the details; I'm hoping to take the Time Series course though so all should soon be revealed.