10-16-2016 11:04 PM
Please find below a description of a project and give me suggestion for analysis and codes. Thank you..:-)
Description of project:
The purpose of this project is to determine the death rate/ survival time of E. coli under a variety of conditions. Field trials were prepared in two locations, LO and CA, during two different seasons, Summer (trial 1) and Fall (trial 2). During each season (trial) at both locations, one field was prepared with three different rates of application to soil, in triplicate. These rates were light, medium, and heavy. Top and core soil samples were collected from each plot and E. coli was enumerated.
Concurrently a microcosm studies was run in the lab. We prepared soil mixture identical to that of the “heavy” application rate in the fields, using soil from both locations, respectively. These microcosms were also prepared in triplicate. The controlled variables in this experiment was temperature, holding the microcosms from both locations at 20°C and 30°C. Essentially, we have three LO microcosms and three CA microcosms at 20°C, and an identical set at 30°C. We monitored the moisture of these microcosms and pH changes over time, along with the populations of E. coli. Soil samples were taken at random from the microcosms.
Questions we want to answer:
Is there a significant difference in E. coli survival between?
What is the best fitting regression for the death rate of E. coli in Trials one and two, locations CA and LO, and application rates Light, Medium, and Heavy?
What is the best fitting regression for the death rate of E. coli in Microcosms 20ºC and 30ºC, Locations CA and LO?
10-17-2016 10:25 AM
SAS is definitely a great tool for your project. Have you started your analysis, or are you new to SAS?
As far as asking "is there a significant difference between...", you can test these statistical hypotheses using SAS. A great reference would be this paper by Art Carpenter.
For regression, I would recommend you start by reading through Regression with SAS, a free eBook from UCLA. They explain many SAS concepts well, and have code examples. There's also a great page on sasCommunity.org about PROC REGRESSION with various model types.
Another book I just found is A Handbook of Statistical Analyses Using SAS. It looks to go through some good examples in detail.
If you put some work into the code and come up with specific questions or issues, I am more than happy to help more.