Dear all,
I have designed a set of experiments to study the effect of different parameters on flocculation of microalgae specie Chlorella vulgaris. The main parameters are OD (optical density) of microalgae cells, pH and amount of flocculant dosage.
The problem is the levels in dosage of flocculant are not constant and they are dependent to level of OD and pH. For example, in OD=1 and acidic pH, the range of dosage varies between 0-100 ppm, though, in OD=5 and acidic pH, this range varies between 0-1000 ppm.
Please let me know how I could incorporate these data in a statistical set to run ANOVA statistical analysis on them.
It seems a multifactor ANOVA. What is your response variable (or dependent variable)?
My work is about studying the effect of different parameters on flocculation of Chlorella vulgaris.
The main parameters (independent variables) for experimental design are optical density of microalgae cells, pH of medium and the dosage of added flocculant. The response parameter (dependent variable) is the flocculation efficiency that shows how much flocculation would be progressed. In my work, I have 4 different levels for OD, 0.4, 1.0, 3.0 and 5.0 and 7 different levels for pH, 3.0, 4.5, 6.0, 7.5, 9.0, 10.5 and 12..
I continue each experiment until the flocculation reaches a maximum and then reduces a minimum with increase in dosage of flocculant at each level of OD and pH, but the problem is the levels for dosage are not identical during all experimental conditions. In cases, the dosage varies between 0:100 ppm and there are others with variation of dosage between 0:1000 ppm.
How could such a configuration be fed into a software for ANOVA analysis?
seems to me that a response surface technique is appropriate. Here go a few readings on this. good luck
Regarding RSM, the problem still remains. How could I add dosage data to the RSM? As I described before, my input variables are OD, pH and dosage. OD has 4 levels and pH has 7 levels but for dosage, the levels and their number are dependent on each level of OD and pH. Thanks very much.
RSM is applicable to your case, see e,g.
Jessy Shaji* and Shital Lodha 2008. Response Surface Methodology for the Optimization of Celecoxib Self-microemulsifying Drug delivery System. Indian J Pharm Sci. 2008 Sep-Oct; 70(5): 585–590. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3038281/
and
Response surface with a simple optimum (which I guess is what you want)
https://support.sas.com/documentation/cdl/en/statug/63347/HTML/default/viewer.htm#statug_rsreg_sect005.htm
Taguchi's methodology is probably inappropriate, in view of the interactios, but you might try some of the possible experimental designs outlined in
http://www.brad.ac.uk/staff/vtoropov/burgeon/thesis_luis/chapter3.pdf
cheers
Dear Dr. Voltolina,
Thanks very much for your kindness and help. Still my problem remains. If you look at the first reference, in table 2, the experimental design is stated. It consists of 3 variables and each has 3 levels. In my case, I don't have exact levels for my dosage variable. On the other hand, the range of variable change for dosage is dependent to the levels of OD and pH. How should I enter such data into a software?
- Badges
- Science topic