Greetings
I sampled form Macroalgae in a coastal area and I have to matrix of data. One is Macroalgae amino acid content and another is environmental data such as temperature, salinity, pH, nitrate, phosphate and etc. Now I am looking for the best statistical method to describe the relationship between these tow matrix. What method is the best one? Is it CCA/RDA, or simple correlation?
Thanks in advance
I think that multiple regression is suitable if you examine the relationship between one dependent variable and several environmental variables. Please refer to
Quinn, G. P., & Keough, M. J. (2002). Experimental design and data analysis for biologists. Cambridge,UK: Cambridge University Press.
If you want to use RDA. I suggest you read the paper as below
FEN GUO, MARTIN J. KAINZ, FRAN SHELDON AND STUART E. BUNN (2015) Spatial variation in periphyton fatty acid composition insubtropical streams. Freshwater Biology. doi:10.1111/fwb.12578.
Hope it helps!
What is the question/goal?
Is there any replication? As in, I sampled 10 locations in the Persian Gulf twice per month for three years. I can now see how the answer changes over location, and correct for seasonal and yearly cycles.
Alternatively, the algae are in a tank with artificial light. That would eliminate some of the variability one should expect from field samples.
I think that multiple regression is suitable if you examine the relationship between one dependent variable and several environmental variables. Please refer to
Quinn, G. P., & Keough, M. J. (2002). Experimental design and data analysis for biologists. Cambridge,UK: Cambridge University Press.
If you want to use RDA. I suggest you read the paper as below
FEN GUO, MARTIN J. KAINZ, FRAN SHELDON AND STUART E. BUNN (2015) Spatial variation in periphyton fatty acid composition insubtropical streams. Freshwater Biology. doi:10.1111/fwb.12578.
Hope it helps!
Multiple linear regression, as mentioned above by Mr. Ouyang, or even nonlinear regression. However, there are several assumptions, diagnostics and preliminary steps to perform first.
Dear Mohammad,
Timothy Ebert's questions are very important. Also, it is essential that you describe what data you have and what they represent of variation in space and time, including the magnitudes of dimension of your data matrices. Until you have done so, any explicit advice of methods to use is likely to be misleading, but afterwards you could presumably get sensible advice by a lot of people.
Best wishes!
Thank You all @Timothy A Ebert, @Xiaoguang Ouyang, @Mohammad Ali Khasawneh and @Rolf Sundberg
Dear Timothy A Ebert
I wanted to eliminate the effect of spatial, so I sampled form one location each midpoint season for a year.
My goal is to find out how environmental factors such as temperature or salinity or even water nitrate can influence on the amino acid content or composition as well. This is a good way to control environmental factors in a Lab conditions, but as you know we have multi-factors in the wild nature and it is not controllable.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis