Hi there Tmsmom
I will start by acknowledging that I am not a microbiologist, nor do I play one on TV!
This sounds like a cool project idea, and I like the use of replication in the design. The inability to work with bacterial cultures may minimize risk, but it makes it pretty difficult to effectively test for removal of bacteria by the filters. And I also think you have a good point about the bacterial test strips being costly.
One thought would be to use a microscope to attempt to look at water samples before vs. after filtration and get a sense of the numbers of microorganisms present. This will work better for protozoa than bacteria, however, but there are protazoa pathogens. The limitation of this approach is that a single or few pathogenic bacteria could be enough to cause illness, and water droplets under microscope will only find organisms when there are a lot present.
Perhaps another idea would be to work with water containing inorganic particles such as silt, and examine the ability of the filters to remove these. This is more of an engineering than a basic science question: how will a filter perform when there is turbid water. Examining such a question, the student would need to think about a number of things that could be educational:
1. How does silt compare to bacteria in its environmental behavior, and what are the strengths and weaknesses of silt as a testing material for water filters?
2. How would one construct a standardized test employing silt?
3. How would she measure silt content in water samples prior to vs. after filtering? (one idea would be to look at the ability of the water to refract light - i.e., turbidity - perhaps there is some kind of turbidity measurement device in the school)
4. What other aspects of filter performance are important besides removing bacteria (e.g., how rapidly the filtration occurs - does the filter clog)?
On the issue of 3 experiments being needed, this pertains to the issue of "replication" and sample size. In short, when comparing the three filters, the only way you know if they have consistent differences is by performing the experiment multiple times on each filter. Then you collect the same measurements for each experiment and can compare the data from the multiple measurements. This deals with the issue that individual experimental results may be variable, and this variability is hard to separate from "real" differences between the filters. Regarding the pre-treatment measurements (i.e., of the pond water before running through the filter), this should also be replicated (i.e., multiple experiments). Perhaps a useful design could be to measure each water sample both before and after the filtration, and look at all the data. If you are able to find some cheap/easy measurements, that makes it easier to do.
Anyways, after you and her are able to make a little more progress on a possible study approach that will be approved, Science Buddies experts can interact with you more to provide additional feedback.
I know this is kind of "long winded" - hopefully some of the brainstorming ideas here will be of some help!
Ben