On the scientific method

dh321 at excite.com dh321 at excite.com
Sun Mar 25 02:42:39 EST 2001

One common problem with school science fair projects on plants is that
students often seem to pick their project and hypothesis without really
searching the available literature first. They may not have access to
research literature but can consult internet sources, encyclopedias,
biology/botany textbooks, science project books, etc. So instead of an
hypothesis that is an "educated guess," it is more of a wild guess. Then
after they do the experiment, they also have no idea how to explain why they
obtained those particular results. Part of that may not be their fault but
was due to their not getting proper guidance in picking a feasible project
in the first place. This newsgroup has discussed problematic science fair
projects before such as music effects on plants and plant growth under
different light colors. I answer questions for the Mad Scientist Network
(www.madsci.org) and we get many questions on those two types of projects
plus questions on other problematic experiments, such as those involving
magnetism effects on plants and watering plants with beverages, such as
milk, soda, tea, and coffee.

Replication is certainly very important, and I have a section in my plant
biology project book saying that replication is essential. More basics of
overall experimental design could be easily introduced into student plant
growth experiments, such as blocking and initial matching of replicates in
each treatment to assure that there are no significant differences among
treatments at the start of the experiment. Another possiblity is having
extra plants to avoid edge effects. Students often also limit their data
collection to plant height, which is often a poor indicator of actual
growth. It is not that difficult to measure plant fresh weight, dry weight,
leaf area, leaf number, etc. and make a photographic record of the plants.
It is much better to have too much data than too little.

Statistical analysis is a tough issue for precollege students. I agree that
statistical analyses typically used by plant researchers, such as ANOVA, are
probably beyond even most high school teachers. Even university researchers
often have statisticians help with their experimental design and data
analysis. I'm not so sure that precollege students should get bogged down
too much with statistics. It would be a big improvement if they had
replications and would place each replication on a graph with the mean so
they can see the actual biological variability. It would not be expecting
too much for them to understand how to use simple statistics such as a
standard deviation.

One thing that might strengthen a presentation on the scientific method
would be go through in detail an actual example of the scientific method in
action, perhaps a graduate student's research describing how they did a
literature search, formulated their hypothesis, designed and executed their
experiment, statistically analysed the results, and drew conclusions. 

David R. Hershey
dh321 at excite.com

Send a cool gift with your E-Card


More information about the Plant-ed mailing list