I think it is important to begin the walk as soon as it gets dark enough for glow worms to start glowing. The reason for this is that we suspect that if there is a swarm of males, they will mate with the available females within a few minutes and as the females stop glowing shortly following the onset of mating, anyone looking for them later would miss them and possibly record small numbers where there had in fact been many more. As a rough guide, I find that glow worms start to appear when the ambient lighting is so low that human colour vision stops working. In the south of England in midsummer this is about 10.30 BST on a clear evening, but in Scotland it would be about an hour later.
The basic recording form on the site is fine for records from the general public, but there is also a more detailed form used by Tim Gardiner of Writtle College which would be preferable for a more scientific study, though you should ideally survey on more occasions if you are aiming for the best results (see below).
Numbers do change from night to night, but usually slowly. If you want to investigate the possibility of male swarming, it would be a good idea to do nightly surveys, at least until mid July, but I know that this is costly in manpower. At the minimum, I'd suggest a weekly survey.
The other aspect is to look for larvae. This is tricky, but they should be active between about April and November, and possibly whenever the weather is mild – maybe over about 10º C. Larvae can sometimes be found by turning over stones and logs during the daytime, and some people have been known to put down small pieces of cut bark to provide hiding places for them. At night, particularly during the summer months, they can be seen glowing briefly on dark nights, more towards the end of the season than the beginning, for unknown reasons. A survey at Tewkesbury attempted to find them using pitfall traps, with limited success. I think the night time hunt in weedy areas is probably the most likely to locate them.
People often do vegetation surveys (usually as part of some degree course) but I think this is a bit pointless. The female glow worms display where there is an open aspect, and they don't care about the vegetation since they can't feed as adults. They often appear on barren ground. A survey of vegetation where larvae are to be seen would be more useful, but the difficulty of finding larvae usually precludes this.
My feeling is that larvae are to be found when and where their mollusc prey are out and about, so any glow worm survey should try to tie in a mollusc survey. As I say, the locations of the glowing females may not be significant. On a site that is geographically restricted, such as a road margin, you will see the females and larvae together. But give them a free rein, and the two locations may be widely different, as Elmhirst suggested. I have heard of a larva being caught in a pitfall trap at least 100 metres away from the nearest glowing female, but in the Tewkesbury survey, they found the larvae mostly where the females were seen.