Science Buddies: "Ask an Expert"

what is the best way to Calculate the organic matter in sediments at the mouths of rivers, and what is the stander method of collecting samples? how many samples are recommended to get accurate results? and what statistical analyzes is recommended to compare 3 different locations (3 different river mouths)?

This is a very interesting project, and your question is a good one. The number of samples required depends on your ability to collect representative samples of sediment, and on the accuracy and precision of your method. I would recommend collecting samples in duplicate or triplicate. If you will be collecting samples from different locations or depths, collect at least two samples at each location.

What method are you going to use to measure the organic matter in the sediment samples? Here is a US-EPA report on various methods for this type analysis, including safety precautions.

http://www.epa.gov/esd/cmb/research/papers/bs116.pdf


Donna Hardy

Donna Hardy, thank you very for your help.

for my project i'm thinking of keeping it simple, i know that there is a standardization on measuring organic matter in sediments, and there are far more complicated ways to get more accurate reading (and expensive ones), but i think the easiest is to due two separate studies, one measure Total carbon = organic carbon by sample combustion to get the percentage of organic matter, (prior to combustion i will remove inorganic carbon by adding the appropriate acids) this will tell me how much but not what kind of organic matter, to solve this issue i will conduct a procreate analysis to Isolate Humic and Fulvic Acids since typically they makes up about 50% of the dissolved organic matter (DOM) in surface water.

Now... i will be compering tree different rivers that lead up to the bay, i still need to know what kind of statistical analysis is the best to compare 3 different locations. this is where i usual get stuck since i never had a good statistics class and i don't really know what to do with my results once i have them.
and without the right statistical analysis to back up my results; well that's 50% of my grade.

Hi,

If the statistical analysis represents 50% of your grade, you need to plan ahead and make sure your data will be suitable for analysis. The grading scale is unusual for a science project, but I think your teacher is trying to make an important point.

What is the overall purpose of your project? What problem are you trying to solve by measuring the humic and fulvic acids and and total organic carbon? It’s important to design your experiments so you will have enough data to analyze. Please let me know more about your project.

At the very minimum, you need to calculate accuracy and precision of your data and you should be able to calculate the standard error. You might use the student’s t-test to determine if there is a significant difference in results between the 3 rivers. Are you familiar with any of these calculations?

Donna Hardy

hello Donna Hardy:
i really appreciate your help, yes i have worked with a student’s t-test; and your right my teacher is trying to make a very important point, this is the 3rd experiment that i have done with this class the class i'm taking is Experimental Designs, and well the word statistics is on the definition of experimental designs that's why its 50% of the grade for each project, the teacher really doesn't gives us any clues just the topic and we are limited only my our imagination, we have to go through all the process of designing it from scratch, i do all the research before starting to plan out what i'm going to, over all we get 3 grades for each experiments, the design is one (and the most important one, because if your design is wrong well everything is wrong), the actual experiment and lab work is another (that includes the treatment of data and results), and the final grade is the report (that is written as if it was to be publish). to be honest its the hardest class i'm taking this semester but its the one have learn the most from, especially since i'm considering a carrier in biology.
once again i wanted to thank you for your help, and i wanted to let you know that i'm almost done with the design and i think i got it covered.

Hi Eduardo,

Thanks for the additional explanation. Let me know when you have your experimental design completed.

Donna

Hi Eduardo,

You have been getting some excellent advice from Donna. I just wanted to note that, depending on your experience with statistics, an Analysis of Variance (ANOVA) is actually the most appropriate statistical test to look at differences among three or more groups. Technically, a t-test is only for comparing two groups. Are you familiar with ANOVA?

Heather

hello Heather;
i did look into a Kruskal-Wallis test. ANOVA for 3 or more independent samples thank you for your help. i'm finish with my proposal the next step would be looking into getting my hands on; porcelain crucibles, analytic balance with a sensitivity of 0.0001 mg. and a electric muffle furnace. (but that's for my teacher to worry about). once again thank you all for your support, this was my first time using "ask an expert" and i hope that one day i could do the same for some one else.