Blather

It has been a complicated time, lately. Emotionally, structurally. Workwise.

On the latter front, there were interviews last week, and there are more next. I tend to talk too much in these things, just blathering on and on and on -- coherently, but in much more detail than anyone should be expected to listen to -- but otherwise they aren't too bad. One of them, sooner or later, will result in employment; I just hope the one that does has geography on its side.

I've also taken, in quick succession, at the behest of an agent and a potential employer, two different online assessments of Java 2 knowledge, which turned out to be, if nothing else, an interesting exercise. One of the features of these online testing setups is that you don't get told the results -- those are delivered to the entity that requested (and presumably paid for) the test, and may or may not be revealed by that entity in the fullness of time -- so I can't say with certainty how I did, although I'm fairly sure I did rather better on one than the other.

The fact that both tests took place on the same day did, however, allow me to compare the two and draw some vague conclusions about this online testing malarkey. Certainly, I think one of the tests was much better structured than the other, and also had more germane content, and will probably give a more accurate appraisal of my skills -- naturally, that's the one I think I did better on. If I was applying such a test to people I was recruiting (heaven forfend), I would have no hesitation in choosing one test supplier over the other, even though the better one is Windows/IE only (hence undertaken on Ian's laptop rather than my own), which on general principle is just evil and rude.

However, the whole process did make me wonder what they think they're trying to get at in doing this.

I have interviewed quite a few people myself, over the years, and even employed some of them. In the process I made various mistakes -- no names, no pack drill -- but I did eventually build up some idea of what I was looking for, and I don't believe these tests would help me find it. On a bad day they'd quite likely keep it away.

This may, of course, be a consequence of the fact that my whole career has been spent in the more disreputable margins of the software industry. Perhaps in the mainstream -- recruiting for investment banks, say -- the sort of pointless hand-wavy nit-picking that is the focus of these tests is what it's all about. But, frankly, I doubt it.

I've mentioned this before, but Joel Spolsky's Guerrilla Guide pretty much hits the nail on the head, it seems to me. For current purposes, I'd especially draw your attention to this little paragraph:

In the end, that is the only question these online tests ask. It's the question that MCSEs or whatever are built on. For any kind of analytical or creative programming, such filters are worse than useless. Of course the details matter -- programming may well be the most pedantic discipline in the world -- but attention to detail is quite a different thing from the regurgitation of domain-specific minutiæ.

Anyway.

The most attractive looming prospect, for my geek self if not my financial situation, is one that no recruiter would consider me for and for which no online test exists. It is only on the horizon because one of the people involved used to work with me; I recruited him, in a way, although really that was just rubber-stamping the decision of someone else (Hi James); later, but still a long time ago, I wrote him some rudimentary physics code on a whiteboard. Also, he apparently has discovered this blog. (Hi Alex.)

Perhaps that just goes to show that the old adage was right all along: it's not what you know, it's who you know.

Bleah. I really should be in bed.
Posted by matt at July 25, 2004 03:48 AM