How do you accurately measure time with a twister-timer? I had boasted that I could do it and so a Year 8 class was duly assembled in order to watch me fail.

My half-metre rule, suspended by threads at each end, would turn gently as it gradually unwound before clattering to the floor in exactly five minutes.

What would we need to know to make such a timepiece? Responses were eager - the number of twists, the length of thread, the weight, the height. And then came the voice that you pray for: “We could change some of these and see how it alters the time.” Brilliant!

So the challenge is set to make a timer that measures three minutes as accurately as possible, but everyone has to work hard because they will all be tested at the end of the lesson.

Moving from group to group, the whole range of approaches is seen. The haphazard let’s-try-it-and-see method is gently redirected while the well-planned experiment is continually encouraged and improved upon.

An understanding of variables and controls can be teased out and predictions, tables of results or graphs can be introduced if you like. Ultimately, the challenge is enjoyable and the title of Most Accurate Timer is an excellent prize.

Teaching practice is daunting, but it is also a time for experimentation and innovation. This lesson, born of the need to try to spice up a Year 8 science course, has been a regular winner for many years.

And what of my own efforts? To my relief, the timer unwound and knocked over empty cans to produce a noisy alarm call that coincided exactly with the end of my five-minute introduction.

Even better, my PGCE tutor saw it all.

Cliff Porter writes about science on www.tes.co.uk