I have a dataset consisting of measurements from 40 locations from 2017 and 2018. These measurements are repeated measurements/matched pairs and it would be simple to test if there is a significant difference in the means between the 2 years by performing a wilcoxons test or a t-test.
In my dataset i have added measurements from 40 additional plots in 2018. These are new measurements and thus unmatched.
My question is: How do i test if there is significant difference between the means (or medians) of the 40 measurements from 2017 and all 80 measurements from 2018 (consisting of both matched and unmatched measurements)?
Use "plotID" as a random factor in a mixed model ("year" is the fixed factor).
I think you use one-way anova. If there is an intervention phase between the requesting measurement and you make the same questionnaire or measurement again I would use the Khi-square. You can do with excel.
You can use one-way ANOVA in the special programme or excel. Significance is tested with 99 % and 95 % accurancy.
- Badges
- Science topic