More science fair advice

Today I gave my 35-minute spiel to the teachers at the K–8 school that I judged at last week.  I did not cover everything in my notes (yesterday’s blog post: Science Fair advice), but got through the first 10 points.  The last two (on graphing and poster design) would probably have taken more time than I had even if I had just covered those two topics.  My discussions on poster design with college seniors and grad students usually take about an hour, and science fair posters aren’t any easier to design—in fact the wider range of audiences, from grade school kids to experts in the subject, probably make them even harder to design well.

A few things came up that I did not have in yesterday’s notes:

1. Good science fair projects take time.  There is a long process in going from a topic to general questions to preliminary experiments to specific questions.  The process needs to be started in October or November and done a little at a time, as a series of littler assignments, rather than a huge push at the end that overwhelms the kids and invites parents to rescue them.  Getting buy-in from the kids and the parents is essential for any sort of homework, but a longer, less intense project is likely to lead to easier relations between parents and teachers.  Kids will almost all need help with time management, but spreading the work out will reduce the conflicts with other demands on their time.
2. Different grade levels should have different expectations of what students need to do, from very scripted group activities at the kindergarten level, to almost independent projects at the 8th grade level.  The transition from teacher-directed to student-directed projects is a gradual one, and each teacher will have to find the level of autonomy that works best with the kids they have to work with.  For example, at third grade, the teacher may decide to limit kids to the physical sciences, because of the greater ease of repeatable measurement and, hence, smaller acceptable sample sizes.
3. Are mentors a good idea? One teacher pointed out that some parents saw having mentors as “cheating”, since the students with mentors were more likely to produce well-crafted, carefully thought-out projects. But mentors can provide much more detailed, specific feedback to students than one teacher can, and are often essential for kids to really learn the topic they are researching. So I’m strongly in favor of encouraging the pairing of kids with mentors.
   In an ideal world, all the students would have mentors, but the supply of mentor time is somewhat limited. For the mentors, the reward is working with kids who are excited by science and interested in topics that the mentor is interested in.  Working with a kid who was just going through the motions and didn’t care about the subject would burn out a mentor and not do the kid much good either.
   Mentors do not need to be scientists or engineers, if the subjects that the students are researching overlap the interests of others.  For example, a kid very interested in music could talk to music teachers about interesting questions that come up in music that can be answered by experiment (for example, what makes an interval consonant or dissonant?).  A kid interested in dance could talk to dance instructors about measuring balance, speed, muscle strength, or ability to follow a rhythm.
4. We talked a lot about the use of preliminary experiments.  For example, if a student builds a trebuchet to play with, the preliminary experiments are usually all over the place: looking at different projectiles, different counterweights, different launch angles, different sling designs, and so forth.  Most students stop at that point and try to write up a poster.  But that is really the starting point for designing an experiment.  What properties were easy to measure? (Distance traveled is easy, time of flight is harder, maximum height of projectile harder still.) What manipulations seemed to have a big effect?  The careful experiment can then be designed to test one particular variable that appeared to be important in the preliminary investigation (for example, just the launch angle) and many repetitions done.
   These preliminary experiments should be done in November or December, so that there is time to design and carry out the “real” experiment, or to change projects, if the measurements needed turn out to be too difficult to do with available equipment.

Related articles

• Science Fair advice (gasstationwithoutpumps.wordpress.com)
• Science fair time again (gasstationwithoutpumps.wordpress.com)
• California State Science Fair judging (gasstationwithoutpumps.wordpress.com)
• Science fair judging (gasstationwithoutpumps.wordpress.com)
• A rant about school science fair projects (geomblog.blogspot.com)
• Anatomy of a Science Fair Project (blogs.scientificamerican.com)