I am working on optimization of extraction procedure for multi-mycotoxin analysis. On this part I want to chose the best extraction solvent among 10 different solvents for the simultaneous extraction of 11 mycotoxins. For this experiment, I first spike the blank sample with the mixed standard of 11 mycotoxins at different concentrations. (as shown in the table attached). The extraction efficiency of the 10 different solvents were evaluated by comparing the recovery (response). In this case I have 11 responses.
On the statistical analysis to check the significant effect of the different solvents on the extraction efficiency of 11 mycotoxins (simultaneous) ,
1. Should I consider the different spiked concentrations and different mycotoxins? If so should I consider them as random factors? Then I would have 2 random factors (mycotoxins/concentration) and one fixed factor (different extraction solvent), So I should use GLM?
2. Or if no need to consider them, should I use one way ANOVA, since I have only one factor of different extraction solvent?
Thank you for your time and consideration.
Hello Yibadatihan
I do not think one way ANOVA would be right as you are actually have two nominal variables - Solvent type and Mycotoxin type. It may be that Solvent 1 recovers one Mycotoxin better compared to other or in the other way around. I personally feel that GLM should be a choice. Furher, within GLM, we should not think about the concentration of mycotoxins (used for spiking) to be a variable; these are just the concentrations with which post-test results would be compared. In my view, solvent type and mycotoxin type would your fixed factors and there would be no need of random factor.
The discussions by RG members would definitely add or correct to my answer.
Maninder
Hello,
as i understand, you have mycotoxins and different conc. of solvents, i think that you may need to run two way ANOVA as you have one independent factor (kind of mycoyoxins) and two dependent factors (the kind of solution and the concentration) so you may need to run ANOVA between groups (kind of mycotoxin) and within groups (between different conc. of each solvent).
i advice you to use SPSS. may be other discussions be beneficial than this.
hope the best
Hello Yibadatihan
I feel that I am at fault at one point that I am not thinking about concentration. If we analyze, it seems to be paired T-test i.e. the concentration before test was X (spiked value) and after recovery was value Y. Does there a difference?
I conclude by saying that probably Repeated Measures ANOVA would be suitable. How to do in SPSS, I do not know
Maninder
Dear Yibadatihan Simayi,
I suggest you to use two way ANOVA in Microsoft Excel. The method is given in the link below:
http://www.real-statistics.com/two-way-anova/two-factor-anova-without-replication/
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA