Is it correct to apply the paired t test for statistical analysis of the results obtained during the reaction time of two different treatments applied to wastewater?
We have applied this statistical method to data of analytical results (along the time) of effluents from two different treatments applied to wastewater to demonstrate whether exist significative differences between both treatments.
If you have paired data, where the matching is over time, a paired samples t-test is fine. If you only have a small number of data points, you should consider a Wilcoxon signed-rank test.
There might be an issue of repeated measures too. Can you be a little more specific?
I agree with the answers given. paired Wilcoxon signed-rank test would be the most conservative analyse to performe. If you need to do further adjustment, you could use a conditional logistic regression (the dependant variable being the binary variable seperating groups).
Parametric methods should be used with normally distributed data and samples taken in a random way. With a small number of data points nonparametric tests are more suitable, but also for small number of samples normality requirement is less restrictive and t test can be used as far as extreme data are not present. Independence or dependency must be considered.
- Badges
- Science topic
- Similar topics
- Biological Science
- Biochemistry
- Biochemical Techniques
- Analytical Chemistry
- pH